Ecologie évolutive d’un genre d’acarien hématophage : approche phylogénétique des délimitations interspécifiques et caractérisation comparative des populations de cinq espèces du genre Dermanyssus (Acari : Mesostigmata) Lise Roy
To cite this version:
Lise Roy. Ecologie évolutive d’un genre d’acarien hématophage : approche phylogénétique des délim- itations interspécifiques et caractérisation comparative des populations de cinq espèces du genre Der- manyssus (Acari : Mesostigmata). Sciences du Vivant [q-bio]. AgroParisTech, 2009. Français. NNT : 2009AGPT0040. pastel-00005531
HAL Id: pastel-00005531 https://pastel.archives-ouvertes.fr/pastel-00005531 Submitted on 26 Feb 2010
HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. N° /__/__/__/__/__/__/__/__/__/__/
THÈSE
pour obtenir le grade de
Docteur
de
l’Institut des Sciences et Industries du Vivant et de l’Environnement (Agro Paris Tech)
Spécialité : Biologie de l’Evolution et Ecologie
présentée et soutenue publiquement par
ROY Lise
le 11 septembre 2009 11 septembre 2009
ECOLOGIE EVOLUTIVE D’UN GENRE D’ACARIEN HEMATOPHAGE :
APPROCHE PHYLOGENETIQUE DES DELIMITATIONS INTERSPECIFIQUES ET CARACTERISATION COMPARATIVE DES POPULATIONS DE CINQ ESPECES DU GENRE DERMANYSSUS (ACARI : MESOSTIGMATA)
Directeur de thèse : Claude Marie CHAUVE Codirecteur de thèse : Thierry BURONFOSSE Travail réalisé : Ecole Nationale Vétérinaire de Lyon, Laboratoire de Parasitologie et Maladies parasitaires, F-69280 Marcy-L’Etoile
Devant le jury :
M. Jacques GUILLOT, PR, Ecole Nationale Vétérinaire de Maisons-Alfort (ENVA).…………...Président M. Mark MARAUN, PD, J.F. Blumenbach Institute of Zoology and Anthropology...…………...Rapporteur Mme Maria NAVAJAS, DR, Institut National de la Recherche Agronomique (INRA)..………... Rapporteur M. Roland ALLEMAND, CR, Centre national de la recherche scientifique (CNRS).……………Examinateur M. Thierry BOURGOIN, PR, Muséum National d’Histoire Naturelle (MNHN)...... ….... ………….Examinateur M. Thierry BURONFOSSE, MC, Ecole Nationale Vétérinaire de Lyon (ENVL)...……………..… Examinateur Mme Claude Marie CHAUVE, PR, Ecole Nationale Vétérinaire de Lyon (ENVL)…...………….. Examinateur
L’Institut des Sciences et Industries du Vivant et de l’Environnement (Agro Paris Tech) est un Grand Etablissement dépendant du Ministère de l’Agriculture et de la Pêche, composé de l’INA PG, de l’ENGREF et de l’ENSIA (décret n° 2006-1592 du 13 décembre 2006)
Résumé
Les acariens microprédateurs du genre Dermanyssus (espèces du groupe gallinae), inféodés aux oiseaux, représentent un modèle pour l'étude d'association lâche particulièrement intéressant : ces arthropodes aptères font partie intégrante du microécosystème du nid (repas de sang aussi rapide que celui du moustique) et leurs hôtes sont ailés. En outre, D. gallinae est une espèce d'importance économique, ce qui rend possible des comparaisons entre colonisation de milieux anthropisés et sauvages. Au début de l'étude, les espèces du groupe gallinae sont très mal délimitées. Les caractères morphologiques utilisés sont variables au sein de l'espèce, voire de la population, très chevauchants entre espèces. Afin de mieux comprendre les exigences écologiques du développement de D. gallinae et d'appréhender ses voies de dissémination, une investigation comparative basée sur des séquences d’ADN entre espèces du groupe gallinae a été adoptée. Un cheminement d'ordre taxinomique a permis de poser les bases nécessaires. Ensuite, l'exploration de certaines caractéristiques écologiques du groupe gallinae en relation avec sa phylogénie (spécificité d'hôte, flexibilité évolutive) a été menée à bien. Une espèce a été décrite, D. apodis, une lignée de D. gallinae constitue aussi une probable espèce inédite et D. longipes regroupe deux entités. Des différences écologiques marquées entre D. gallinae et les autres espèces semblent résulter non seulement de l'activité humaine, mais aussi de caractéristiques intrinsèques. Aujourd'hui, le rôle des flux commerciaux dans la dispersion de D. gallinae en élevage de pondeuses s'avère primordial, au moins en France, celui des oiseaux sauvages presque nul.
Mots-clés Système hôte-parasite, microprédateur, phylogénie, espèce, population, microhabitat, Dermanyssus, mésostigmate, acarien
1 Evolutionary ecology of a hematophagous mite genus: Phylogeny-based approach for interspecific delineations and comparative characterization of populations in five species of the genus Dermanyssus (Acari: Mesostigmata)
Abstract Micropredator species of Dermanyssus (Moss'gallinae-group), which parasitize birds, represent an interesting model for the study of loose associations. Thus, these unwinged arthropods do not stay on host (blood meal as quick as mosquitoe's), are part of the nest's microecosystem and their hosts are winged. Moreover, micropredator Dermanyssus include at least one species of economic importance in fowl farms, D. gallinae (the Poultry Red Mite), which enables the comparison between species restricted to wild avifauna and synanthropic species. At the beginning of the study, micropredator species are not clearly delimited. Most of species specific morphological characters are variable within species, in some cases within population, and are overlapping between species. In the aim to investigate the ecological needs for proliferation in D. gallinae and its ways for dispersal, a DNA-based comparative analysis involving this species and its close relatives has been performed. The first section consists of the clearing of the taxonomy and species delineations. In the second section, ecological and intrinsic data (host specificity, flexibility) are compared between species of the gallinae-group. One species has been described (D. apodis), one lineage within D. gallinae seems to represent a cryptic species and D. longipes currently groups two different entities. Important ecological differences between D. gallinae and other species seem to result not only from human activities, but also from intrinsic characteristics. Currently, the role of trade flows in D. gallinae’s spread in layer farms appears to be essential, at least in France, as opposed to the role of wild birds (nearly nul).
Key words Host-parasite system, micropredator, phylogeny, species, population, microhabitat, Dermanyssus, Mesostigmata, Acari
2 Remerciements
Je tiens à remercier du fond du cœur mes deux directeurs de thèse, à qui je dois, c’est certain, beaucoup plus qu’il n’est commun de devoir à ses directeurs de thèse.
Merci à Claude Chauve, pour le soutien et la confiance qu’elle m’a admirablement prodigués tout au long de cette thèse. Je ne sais comment exprimer ma reconnaissance et mon respect pour son courage et sa loyauté remarquables.
Merci à Thierry Buronfosse, pour le soutien, la confiance et la patience dont il m’a fait bénéficier au cours de cette thèse ! Nos continuelles oppositions, contradictions et confrontations ont sûrement été la clé du présent travail. Plus qu'un encadrant, c’est un collègue, voire un ami, que je crois avoir trouvé en lui, qui m'a soutenue et aidée au-delà de toute espérance. Et il a réussi à former une « prof. de Lettres » à la biologie moléculaire !
J'aimerais que cette thèse puisse constituer un remerciement suffisant à l’appui si solide, si bénéfique et sans cesse renouvelé dont ils ont tous deux fait preuve.
J'adresse mes vifs remerciements à Maria Navajas et Mark Maraun pour avoir accepté d’endosser l’important rôle de rapporteur.
Je remercie cordialement Roland Allemand, Thierry Bourgoin et Jacques Guillot pour avoir accepté de prendre part au jury de cette thèse. Je remercie aussi les deux premiers pour leur suivi dans le cadre du comité de pilotage.
Merci à l’ensemble de mes collègues de travail, passés ou présents, en particulier mes collègues de laboratoire qui travaillent ou ont travaillé sur Dermanyssus : Claire Valiente Moro, Sophie Desloire, Marie-Thérèse Poirel. Une pensée émue à Marie-Claude Reynaud.
Merci à mes valeureux stagiaires, qui ont subi mon obsession du genre Dermanyssus au long de stages de durée variable, ont, pour la plupart, fourni un travail particulièrement satisfaisant et ont ainsi pleinement participé au travail de recherche que voici : Robin Bellon, Sandrine Bonnet, Jean Filippi, Mathieu Galès, Mehdi Gharbi, Nina Guichard, Guillaume Lallemand, Sophie Merlin, Sabrina El Ouartiti, Anthony Piron, Marie Rigaux.
Merci à l’ensemble des personnes qui ont contribué à l’obtention des indispensables échantillons biologiques : Olivier et Nathalie Chovet, le Centre de Soins aux Oiseaux Sauvages du Lyonnais, notamment Pascal Tavernier, Mathilde, l’Ecopôle du Forez, notamment François Boléat, Claire Brucy, Pascal, …., la Réserve du Romelëare, le Centre de Recherche sur la Biologie et les Populations d’Oiseaux, notamment Gérard Gory, Olivier Caparros, Laurent Brucy, Laurent Carrier, Nicolas Vincent-Martin, Yves Beauvallet, Olivier Dehorter, Franck Salmon, Rolf Wahl, J. Girard-Claudon,… En particulier, merci à Olivier Caparros pour m’avoir accueillie au cours de ses passionnantes séances de baguage d’oiseaux sauvages, et m’avoir instruite quant à la biologie et les traits d’histoire de vie de ces animaux fascinants.
3 Merci à l’Association Ornithologique Rhodanienne, en particulier M. André Marchal et M. Marc Serre, pour m’avoir permis d’obtenir des échantillons provenant d’élevages de petits oiseaux de compagnie et/ou de compétition. J’adresse également mes sincères remerciements à M. Jean- Jacques Sage, président du Club Français du Montauban et M. S. Calabro, président de l’Amicale Ornitholoqique Becs Crochus Centre Est.
Merci à Christine Basset pour avoir pris le temps de m’accompagner et m’avoir ainsi si souvent permis d’entrer dans tous ces élevages de volailles en Bresse, pour sa sympathie et sa bonne volonté permanentes. Je remercie également le Dr Patrick Chabrol, Jérome Arnal, … , pour leur aide aux prélèvements dans les élevages de pondeuses.
Merci à Sophie Lubac pour son aide dans la collecte d’acariens provenant d’élevages de volailles et les informations concernant la filière pondeuse. Son implication, son soutien et son dynamisme ont fortement contribué à une bonne partie de l’étude.
Merci au PEP Avicole (Pôle d’Expérimentation et de Progrès, Région Rhône-Alpes) pour l’appui financier sans lequel l’étude, entamée au cours de cette thèse et prolongée au-delà, des voies de dissémination de l’acarien utilisant les outils de la génétique des populations n’aurait pas pu voir le jour.
Merci à Ashley P.G. Dowling pour sa remarquable participation aux travaux de phylogénie inclus dans cette thèse. Son ouverture d'esprit, sa réactivité et son enthousiasme ont accompagné et soutenu très efficacement l'élaboration de ce travail.
Merci à Jean-Charles Bouvier et Claire Lavigne pour leur précieuse et sympathique collaboration. En espérant qu'elle puisse être prolongée au-delà de ce travail de thèse.
Merci à Maurice Sabelis et Iza Lesna pour leur précieuse collaboration et pour m'avoir fait connaître leur passion de l'écologie des acariens.
Merci à l’ensemble de l’équipe de l’EURAAC pour son soutien précieux et l’invitation au congrès de l’été 2008 à Montpellier.
Merci à Joao Sollari Lopes pour ce début de collaboration prometteur dans le domaine de la génétique des populations.
4 Sommaire Résumé ______1 Mots-clés ______1 Abstract ______2 Key words ______2 Remerciements ______3 Sommaire ______5 Liste des publications constituant la thèse______8 Table des figures______9 Table des tables______12 1 Avant-propos______14 1.1 Ecologie évolutive et démarche cladiste______14 1.2 Systèmes hôtes-parasites ______15 a - Cas des ectoparasites ______15 b - Microprédation et faible spécificité d’hôte, en lien avec l’essaimage ______16 c - Difficultés dans l’appréhension des associations lâches ______17 d - Intérêt de l’étude de systèmes microprédateur aptère / macroproie ______18 1.3 Modèles microprédateur aptère / oiseau ______19 2 Introduction ______21 2.1 Contexte : le genre Dermanyssus et le groupe gallinae ______21 a - Une espèce d’importance économique dans un genre méconnu______23 b - Impact direct sur l’hôte ______24 c - Distribution ______25 2.1.c.1 Spectre d’hôte______25 2.1.c.2 Répartition géographique ______25 d - Reproduction ______25 e - Particularités de la biologie des espèces du genre Dermanyssus en lien avec les difficultés de traitement rencontrées en élevage______28 2.1.e.1 Variabilité du temps de génération ______28 2.1.e.2 Etroite relation avec le microenvironnement (nid, litière)______30 2.2 Problématique ______31 2.3 Objectifs ______33 2.4 Aperçu sommaire de l'étude : un débroussaillage en deux étapes ______33 a - Structure du texte______33 b - Première étape (§4) : clarification de l’identité spécifique ______33 c - Seconde étape (§5): exploration écologique ______34 3 Grandes lignes de la méthodologie adoptée______35 3.1 Matériel biologique : stratégie d’échantillonnage pour une représentation d’habitats variés______35 3.2 Marqueurs développés : utilisation concomitante de données morphologiques et moléculaires 35 5 3.3 Outils de la phylogénie et de la génétique des populations ______36 4 Taxinomie dans le genre Dermanyssus ______38 4.1 Synthèse historique : publication I______38 a - Présentation ______38 4.1.a.1 Objectifs ______39 4.1.a.2 Principaux résultats ______39 b - Remarques sur la publication I ______39 4.1.b.1 Nombre d’espèce augmenté (2008 et 2009) ______39 c - Publication I______40 d - Erratum attenant à la publication I ______52 4.2 Evaluation des caractères morphologiques discriminants entre espèces : publication II 53 a - Présentation ______53 4.2.a.1 Objectifs ______53 4.2.a.2 Matériel et méthodes ______53 4.2.a.3 Principaux résultats ______53 b - Remarques sur la publication II ______54 4.2.b.1 Des caractères réhabilités a posteriori______54 c - Publication II ______55 4.3 Délimitation des espèces par une approche complémentaire (« total evidence approach »): publication III ______65 a - Présentation ______65 4.3.a.1 Objectifs ______65 4.3.a.2 Matériel et méthodes ______65 4.3.a.3 Principaux résultats ______66 b - Remarques sur la publication III ______66 4.3.b.1 D. longipes : deux lignées différentes ?______66 4.3.b.2 Un marqueur moléculaire abandonné : EF1-D ______67 4.3.b.3 Liponyssoides : genre introuvable ?______67 c - Publication III ______69 5 Ecologie comparée des cinq espèces françaises du genre Dermanyssus ______94 5.1 Spécificité d’hôte chez cinq espèces du genre Dermanyssus : publication IV ____ 94 a - Présentation ______94 5.1.a.1 Objectifs ______94 5.1.a.2 Matériel et méthodes ______94 5.1.a.3 Principaux résultats ______94 b - Remarques sur la publication IV______96 5.1.b.1 Données complémentaires sur la spécificité d’hôte chez D. hirundinis en France ______96 c - Publication IV ______97 5.2 Diversité génétique et flux de populations chez quelques espèces du groupe gallinae : publication V ______125 a - Présentation ______125 5.2.a.1 Objectifs ______125 5.2.a.2 Matériel et méthodes ______126 5.2.a.3 Principaux résultats ______126 b - Publication V (soumise) ______131 5.3 Arthropodofaune de nids d’oiseaux en agroécosystème et implication des Dermanyssoidea hématophages : publication VI ______211 a - Présentation ______211 5.3.a.1 Objectifs ______211 5.3.a.2 Matériel et méthodes ______212 5.3.a.3 Principaux résultats ______212 b - Publication VI (soumise)______214 6 6 Discussion______239 6.1 Relations phylogénétiques______239 a - Des lacunes ______239 b - Topologies bifides ou en escalier ? ______239 6.2 Etat de la taxinomie du genre Dermanyssus ______240 6.3 Deux marqueurs complémentaires pour l’exploration intraspécifique ______241 6.4 Patterns écologiques révélés ______242 a - Spécificité d’hôte______242 6.4.a.1 Paramètres écologiques ______242 6.4.a.2 Paramètres intrinsèques ______243 b - Transition sauvage-synanthrope : hybridation et radiation chez D. gallinae ______244 c - Structure de populations ______245 6.5 Du caractère invasif de D. gallinae et d’une espèce peut-être concurrente _____ 246 7 Conclusions et perspectives ______248 7.1 Conclusions sommaires ______248 a - Nouveautés taxinomiques ______248 b - Nouveautés écologiques ______248 7.2 Perspectives ______250 a - Exploration des flux de populations au sein de l’espèce D. gallinae ______250 b - Espèces cryptiques ______250 c - Analyse moléculaire et morphologique de l’ensemble du genre Dermanyssus ______250 d - Cophylogenèse au niveau population ______251 e - Investigation de la situation symétrique entre D. gallinae et O. sylviarum en France et aux Etats- Unis ______251 f - Comparaison des valeurs de polymorphisme de séquences d’ADN nucléaires et mitochondriales entre espèces de microprédateurs aptères ______251 8 Lexique ______252 9 Références bibliographiques ______252 10 Annexes______259 10.1 Annexe 1 : apercu de la classification des Mesostigmata selon Hallan 2005 ____ 259 10.2 Annexe 2 : publications connexes______254 a - Généralités sur D. gallinae et les problèmes engendrés en élevage ______254 b - Chimiorésistance vis-à-vis d’organophosphorés chez D. gallinae ______294 c - Participation à une publication sur une problématique de lutte biologique contre D. gallinae ___ 295 10.3 Annexe 3 : numéros d'accès des séquences obtenues d'EF1-alpha et amorces correspondantes ______296
7 Liste des publications constituant la thèse La présente thèse repose sur les six articles dont voici les références. Chacune des publications apparaît, dans le document, isolée par un intercalaire de couleur.
Publication I : Roy L., Chauve C. M. (2007) Historical review of the genus Dermanyssus (Acari: Mesostigmata: Dermanyssidae). Parasite 14:87-100 (pp. 34-48 du présent document)
Publication II : Roy L., Chauve C. M. (in press) The genus Dermanyssus Dugès, 1834 (Acari : Mesostigmata : Dermanyssidae): species definition. Proceeding of the International Congress of Acarology, Amsterdam, August 2006, (publication prévue dans Experimental and Applied Acarology). (pp. 55-64)
Publication III : Roy L., Dowling A.P.G., Chauve C.M. and Buronfosse T. (2009) Delimiting species boundaries within Dermanyssus Dugès, 1834 (Acari: Mesostigmata) using a total evidence approach. Molecular Phylogenetics and Evolution 50:3:446-470 (pp. 69-93)
Publication IV : Roy L., Dowling, A.P.G., Chauve, C.M. Lesna I., Sabelis M.W. and Buronfosse, T. (2009) Molecular phylogenetic assessment of host range in five Dermanyssus species. Experimental and Applied Acarology 48: 115-142 (pp. 97-124)
Publication V : Roy L., Lopes J.S., Dowling A.P.G., Chauve C.M., Buronfosse T. (soumis). Dermanyssus gallinae (Acari: Mesostigmata) possesses characteristics of an invasive species, compared to four other Dermanyssus species. (soumission le 17 juin 09 à Molecular Biology and Evolution) (pp. 131-210)
Publication VI : Roy L., Bouvier J.C., Lavigne C., Galès M., Chauve C.M., Buronfosse T. (soumis) Arthropodofauna in bird nests as an indicator for agricultural practices' impact in pear and apple orchards. (soumission le 26 juin 09 à Ecological indicators) (pp. 214-241)
8 Table des figures
Corps du texte
Figure 1. Aperçu de la classification et de la composition du genre Dermanyssus au début de l’étude, selon Moss (1978). ……………………………………………………………………………………... 21
Figure 2. Cycle de vie des espèces du groupe gallinae.………………………………………………...………....23
Figure 3. Caractères permettant la séparation des stades/sexes chez D. gallinae. ……………………26-27
Figure 4. Agrégat de D. gallinae amassés au fond d’un angle de sachet plastique, en pleine lumière. …………………………………………………………………………………………………………………30
Figure 5. Chaetotaxie des pattes chez les acariens mésostigmates. ………………………………..…………54
Figure 6. Reconstruction phylogénétique avec indication d’hôte et de milieu d’échantillonnage intégrant 73 isolats du groupe gallinae sur la base de la mt-Co1.…………………….…….………95
Figure 7. Topologie finale retenue (consensus strict).……………………………………………………….…..128
Figure 8. Aperçu de la composition du genre Dermanyssus à l’issue de la présente étude.……...…241
Publication I (les numéros de page renvoient à la numérotation de l'article lui-même)
Figure 1. Distribution map of non-gallinae species of Dermanyssus.…………………..……………………88
Figure 2. Mouth parts of D. gallinae……………………………………………………………………………………89
Publication II
Figure 1. Dorsal shield of 10 from 20 randomly selected adult females of a cultured in lab population of D. gallinae…...……………………………...……………………....…...…………………………61
Figure 2. Relative length of peritrema according to position of coxae. ………..……………………………62
Figure 3. Peritrema (scanning electron microscope) in D. gallinae. …………………………………………63
9 Publication III (les numéros de page renvoient à la numérotation de l'article lui-même)
Figure 1. Strict consensus tree of 12 most parsimonious trees (L = 129, CI = 0.4264, RI = 0.6085) using matrix of 46 morphological characters.. ……………………………………………………………451
Figure 2. Molecular combined analysis using 1570 bp from cytochrome oxidase subunit I, rRNA 16S and rRNA 18S-28S, including ITS1, 5.8S and ITS2. ……………………………………………453
Figure 3. Total evidence analyses using 46 morphological characters and 1570 bp from cytochrome oxidase subunit I, rRNA 16S and rRNA 18S-28S, including ITS1, 5.8S and ITS 2.……………………………………………………………………………………………………………………454
Figure 4. Distribution of percentages of pairwise divergence among populations of the eight OTUs used in molecular analyses.. ……………………………………………………………………………………455
Figure 5. Bayesian analysis of the COI matrix, excluding 3rd positions of codons. …………………455
Figure 6. D. apodis n. sp..…………………………………………………………………………………………………459
Publication IV (les numéros de page renvoient à la numérotation de l'article lui-même)
Figure 1. Percentage of occurrence of Dermanyssus in nests of the five bird groups under test. …126
Figure 2. Maximum Parsimony analysis (mt-Co1). …………..……….…………………………………………128
Figure 3. Bayesian analysis (mt-Co1). ..………………………………………………………………………………129
Publication V
Figure 1. Overview of the distribution of variable elements along the studied Tropomyosin sequence (mutation points and insertion/deletion sites).………………………………………………206
Figure 2. Haplotypes distribution according to the six ecological categories ……..………………..…206
Figure 3. Amount of “pure synapomorphies” ………………….……………………………………………..207
Figure 4. Haplotypic phylogenies obtained with mt-Co1 and Tropomyosin sequences.………...207-210
Figure 5. Most supported topologies for the population genetics analysis using ABC methods.....210
10 Publication VI
Figure 1. Distribution of arthropod-rich, arthropod-poor and arthropod-free nests according to control conditions……………………………………………………………………………………………...……236
Figure 2. Percentage of occurrence of arthropoda groups as defined in material and methods. ..…236
Figure 3. Average number of arthropodan groups in arthropod-rich or arthropod-poor nests……...237
Figure 4. Number of nests containing one or several species belonging to the two hematophagous mite families isolated in present study according to control management. ………….…………237
Figure 5. Phylogenetic topology involving individuals of D. gallinae and O. sylviarum isolated in the framework of present study (sampled JBOn), along with some individuals obtained from different environments and places. …………………………………………………………...………238
11 Table des tables
Publication I (les numéros de page renvoient à la numérotation de l'article lui-même) Table 1. Species included or previously included in Dermanyssus listed in chronological order with their present position/status……………………………………………………………………………90-91
Table 2. List of species currently included in Dermanyssus and their known host species..……93-95
Publication II Table 1. Host diversity for Dermanyssus species (from literature data). .….………………………………60
Publication III (les numéros de page renvoient à la numérotation de l'article lui-même) Table 1. Taxonomic sampling and EMBL accession numbers for each sequence……...……………449
Table 2. Primer sequences and key parameters for PCR conditions. ………………………………………450
Table 3. Bootstrap (%) and relative Bremer support for monophyly of species with several populations and other groups in the three single genes analyses (maximum parsimony and bayesian). ……………………………………………………………………………………………………………456
Publication IV (les numéros de page renvoient à la numérotation de l'article lui-même) Table 1. Description of restricted study areas (cf. § Materiel and Methods). ………………………118-119
Table 2. Primer sequences…………………………………………………………………………………………………120
Table 3. Some data on development of 3 Dermanyssus species on canaries, obtained from long- time bioassays (referred to in text as comp1 and comp2). ……...……………………………………121
Table 4. Engorgment and development of 3 Dermanyssus species compared using three host species and short-time bioassays (referred to in text as comp3 to comp7). ………………123-124
Table 5. Number of nests analysed per bird family and occurrence of genus Dermanyssus based and on present field data and on literature. ………………………………………………………………125
12 Publication V
Table 1. Primer sequences.…..……………………………………………………………………………………………166
Table 2. Evaluation of the three-times rule for species and other entities of Dermanyssus under test………………………………………………………………………………………………………………………166
Table 3. Pairwise Fst estimates between D. gallinae focused isolates and between specific datasets, corresponding P values and associated Nm for both nuclear and mitochondrial loci….…167
Table 4. Computer simulations of coalescent process (DnaSP v5) given the number of segregating sites S, assuming an intermediate level of recombination R=10 for Tropomyosin amplicon and no recombination for COI amplicon (confidence interval =95%).…………………………..168
Table 5. Estimates of modes and 95% credible intervals for the considered demographic parameter for the D. gallinae groups with different host types (case I, popABC)…………………………..169
Table 6. Estimates of modes and 95% credible intervals for the considered demographic parameter for the D. gallinae populations in different geographical locations (case II, popABC). …..169
Table 7. Estimates of modes and 95% credible intervals for the considered demographic parameter for the Dermanyssus species (case III, popABC). ………………………………………………………170
Table 8. Genotypic and heterozygosity variability in focused isolates for Tropomyosin exon n, intron n and exon n+1. ……………………………………………………………………………………………170
Publication VI
Table 1. Species detected in the four focused recurrent primary groups. …………………………………235
Table 2. Number of species within Coleoptera and Mesostigmata in arthropod-rich and arthropod- poor nests sampled in 2007. ……………………………………………………………………………………235
13
1 Avant-propos
1.1 Ecologie évolutive et démarche cladiste L’écologie évolutive, s’étend dans le champ de deux disciplines, la biologie évolutive et l’écologie. Elle étudie à la fois les influences historiques et contemporaines sur les patrons de variabilité observés et ce à tous les niveaux, depuis l’individu jusqu’aux communautés d’espèces ou grands groupes taxinomiques. Par excellence transdisciplinaire, elle fait usage d’outils variés, associe modélisations mathématiques de problèmes biologiques et approche expérimentale et représente une approche intégrée des interactions entre les gènes, les individus, les populations et l’environnement. S’attachant à prendre en compte les contingences historiques pour tester des hypothèses adaptatives, elle vise à apporter des éléments clefs pour une meilleure compréhension de l’importance des processus à l’origine des patrons de variabilité observés à différents niveaux dans les systèmes biologiques. La cladistique (du grec klados, branche) mettant à profit les analyses phylogénétiques pour l'appréhension de patrons écologiques, est une des facettes de l'écologie évolutive moderne. Reconstruire une histoire des relations phylogénétiques sur la base de données morphologiques, moléculaires, biochimiques, etc. permet, a posteriori, d'établir le scénario évolutif de tel ou tel aspect plus ou moins intimement lié aux taxa cibles. Quelle que soit la nature des caractères utilisés, le recours aux algorithmes phylogénétiques, qu'ils reposent sur le critère du maximum de parcimonie, du maximum de vraisemblance ou utilisent les méthodes bayésiennes, permet une appréhension raisonnée, et la plus objective possible, des relations évolutives entre lesdits taxa, indépendamment du questionnement initial. Ainsi, a posteriori, au vu des topologies retenues par ces méthodes objectives, des motifs évolutifs peuvent émerger, après un travail de mise en relation d’informations extérieures à l'arborescence avec les relations figurées par les branches à différents niveaux dans l’arbre retenu. Ces informations peuvent être des caractères intrinsèques des taxa mêmes, tels des caractères morphologiques, des mutations sur une séquence d’ADN donnée, des traits d’histoire de vie. Dans ce cas, on établit un scénario évolutif desdits taxa. Par exemple, l’observation de l’arbre retenu avec examen des états de certains caractères morphologiques aux différents nœuds de l'arbre permet d'observer leur évolution depuis l'ancêtre commun (racine), jusqu'aux taxa étudiés (feuilles), en passant par les ancêtres communs internes (nœuds). Des informations plus indirectement liées aux taxa cibles peuvent être corrélées aux arbres obtenus, telles le type d’habitat, la localisation géographique, etc. Des corrélations entre information historique et constat écologique contemporain plus ou moins attendues se dessinent. Les taxa groupés dans un clade donné peuvent s’avérer inféodés à un type d’habitat commun, différent de celui des autres clades, et signer un fait évolutif de nature écologique particulier. Ou encore, une absence de corrélation peut être mise à jour.
La démarche se déroule en deux étapes majeures : (1) la collecte et le traitement phylogénétique des données, (2) la mise en relation des données obtenues sur la base d’une matrice rigide avec des données différentes, plus ou moins dispersées. La première étape, à visée fondamentalement objective, est gérée par des algorithmes. La seconde étape, en revanche, n'est pas automatisable, tout au moins pas complètement. Elle repose sur la capacité à mettre en lien, en réseau, des informations éparses, qu'a priori rien ne lie nécessairement, et des topologies phylogénétiques. Convergence, réversion, dérivation successive d'états multiples apparaissent grâce à l’examen de la disposition des informations disponibles en fonction des clades sur
14 l’arborescence retenue. L'effort d’interconnexion entre les données écologiques, géographiques ou comportementales et l’histoire mise à jour par les relations de parenté permet la révélation de patrons de variabilité. 1.2 Systèmes hôtes-parasites L'histoire d'associations biologiques peut être observée de cette manière, telles des associations proie-prédateur, hôte-parasite, hôte-symbiote, … Les systèmes hôte-parasite en particulier constituent des modèles d’étude de l’écologie évolutive exemplaires. En effet, l’association plus ou moins étroite entre le parasite et son hôte, induisant des contraintes évolutives importantes, offre un vaste terrain pour l’étude des phénomènes adaptatifs. La conjonction de particularités intrinsèques de l’hôte comme du parasite et des caractéristiques écologiques de l’hôte (habitat, habitudes) et du parasite permet à ce système d’exploitation de durer. Ainsi Morand et Sorci (1988) ont-ils démontré, en comparant des nématodes parasites avec des nématodes libres, que, dans la plupart des cas, pour les parasites, l’évolution des traits d’histoire de vie est directement dépendante de caractéristiques de l’hôte.
a - Cas des ectoparasites Les ectoparasites en particulier offrent une diversité dans le degré d’association avec l’hôte tout à fait remarquable (spécificité d’hôte, relations avec environnement de l’hôte). Certains sont aussi intimement liés à leur hôte que la plupart des endoparasites, montrant un haut niveau de spécialisation (ex. les poux mallophages parasitant des rongeurs du genre Geomys ; Page et Hafner 1996). D’autres, moins spécifiques, manifestent des liens un peu plus lâches (ex. les puces du genre Pulex, parasitant des mammifères aussi divers que l’homme, le renard, le blaireau, le hérisson…). Les degrés de spécificité sont variables entre groupes de haut niveau taxinomique, mais aussi parfois entre espèces proches (Desdevises et al. 2002, Price et al. 2003).
Les habitudes et traits d’histoire de vie sont ainsi très divers parmi les ectoparasites, certains accomplissant toutes les étapes de leur développement directement sur l’hôte, d’autres passant certains de leurs stades sur l’hôte, les autres dans l’environnement. Les poux mallophages, ectoparasites d’oiseaux ou de mammifères, les trématodes monogènes, ectoparasites de poissons, sont des exemples d’ectoparasites au développement complet sur l’hôte. Chez la plupart des puces (Siphonaptera : Pulicidae, Ceratophyllidae), en revanche, le stade adulte demeure sur l’hôte, mais les œufs sont généralement pondus dans le milieu extérieur où les larves et nymphes se dévelopent ensuite. Les femelles adultes des puces chiques (Siphonaptera : Tungidae) pondent directement sur l’hôte, mais les œufs tombent au sol. Chez certains arthropodes hématophages, des liens très lâches avec l’hôte et des habitudes alimentaires non strictement parasites tout au long du cycle de vie les rendent difficilement qualifiables de parasites. Chez les moustiques, par exemple, les femelles adultes sont hématophages, tandis que les mâles et les stades juvéniles ont des habitudes complètement différentes (se nourrissant respectivement de nectar et autres liquides sucrés, et de microorganismes aquatiques). Ces femelles adultes sont des microprédateurs, prédateurs qui ne prélèvent qu’une petite portion de tissu de leur hôte. Les arthropodes hématophages pourraient ainsi être classés soit parmi les ectoparasites typiques, soit parmi les microprédateurs. Une telle distinction doit-elle être associée à la constance de leur statut de consommateur tout au long leur vie (hématophage/non hématophage) ou à la proportion de leur cycle passée directement sur l’hôte ? Quoi qu’il en soit, comme chaque fois que l’on cherche à classer quelque chose dans des catégories, des cas limites viennent brouiller les frontières. Les puces, dont les œufs, larves et nymphes se développent dans l’environnement, évoquent fortement des
15 microprédateurs, mais elles demeurent sur l’hôte au stade adulte. Les punaises de lit tendraient à être classées parmi les microprédateurs, car elles ne requièrent pas plus de temps que les femelles de moustique pour prélever leur repas de sang, et, elles aussi, quittent leur hôte immédiatement après, mais tous leurs stades sont hématophages. Et il en est de même pour les dermanysses ou poux rouges des poules (Acari : Mesostigmata : Dermanyssus), ainsi que pour les tiques molles (Acari : Ixodida : Argasidae). Ectoparasites typiques ou microprédateurs ? Une réflexion quant à ces catégories présente un intérêt non négligeable pour l’interprétation écologique des histoires évolutives d’ectoparasites. Mais plutôt qu’à une durée de contact, donnée continue et par trop relative, Kuris et Lafferty (2000) attribuent une importance au nombre d’individus hôtes parasités/prédatés par stade chez le parasite. En effet, dans le cadre d’une réflexion quant aux catégories de consommateurs en général, ils mettent en avant la corrélation entre le degré d’association avec l’hôte et l’attachement de l’individu parasite à l’individu hôte. Cet attachement à l’individu hôte/proie est en quelque sorte inversement proportionnel au nombre d’individus ponctionnés par un seul parasite à un stade donné. Les femelles adultes du moustique, par exemple, peuvent piquer plusieurs individus hôtes différents, et ne demeurent pas attachées à un seul. Cela les rapproche des prédateurs, qui se nourrissent successivement de différentes proies. Si l’on compare des arthropodes hématophages, les microprédateurs sont par excellence plus indépendants que les ectoparasites typiques vis-à-vis de leur hôte et beaucoup plus impliqués dans les environnements extérieurs à l’hôte.
b - Microprédation et faible spécificité d’hôte, en lien avec l’essaimage L’indifférence augmentée du microprédateur quant à l’identité individuelle de sa macroproie oriente par excellence ce type de consommateur vers un plus large spectre d’hôtes. Price (1975) démontre l’extrême réduction du spectre des espèces consommées chez les insectes parasites (tant végétaux qu’animaux) si l’on compare aux insectes prédateurs. Intermédiaire entre l’ectoparasite typique et le prédateur, le microprédateur hématophage est indifférent ou presque à l’individu qu’il ponctionne, et ainsi plus à même de changer d’espèce d’hôte que l’ectoparasite typique. Et sa mobilité propre tend par conséquent à jouer un rôle dans l’ampleur du spectre de ses hôtes. Chez les parasites typiques, les transferts d’hôte à hôte au sein de la même espèce sont fréquents (contagion). Certes, certains cycles parasitaires impliquent des hôtes intermédiaires (cycles hétéroxènes, ex. la grande douve Fasciola hepatica), mais, si ces hôtes peuvent être très distants phylogénétiquement entre eux (ex. mammifère – mollusque dans le cas de la grande douve), ils appartiennent à un système écologique fermé. Dans les systèmes microprédateur - macroproie, les frontières écologiques sont par excellence plus ténues. La définition des microprédateurs par Kuris et Lafferty (2000) s’applique là encore non seulement aux consommateurs se nourrissant de tissus animaux, mais aussi aux consommateurs se nourrissant de tissus végétaux. Chez de nombreux pucerons, un cycle complexe alterne générations à reproduction parthénogénétique aptères avec générations à reproduction sexuée aptes à essaimer (sur un individu de la même espèce de plante hôte ou d’une autre espèce), des générations parthénogénétiques pouvant aussi être ailées et participer à l’essaimage chez certaines espèces. Chez Myzus persicae par exemple, un puceron d’une génération parthénogénétique aptère, demeurant sur une seule et même plante hôte (individu) ponctionne les liquides d’une seul individu tout au long de sa vie et se comporte en parasite typique. Les individus ailés (virginipares ou sexupares) sont voués à l’essaimage et présentent les caractéristiques des microprédateurs. Ce sont ces microprédateurs qui permettent la dissémination des populations.
Chez les hématophages, les femelles adultes ont en général besoin de repas de sang pour la maturation de leurs œufs (Prasad 1987). Ce qui semble différencier en premier lieu la femelle adulte microprédatrice des moustiques (Diptera : Culicidae) par exemple de celle d’ectoparasites 16 typiques, comme les poux anoploures (« poux piqueurs » ; ex. Pediculus hominis), c’est la capacité pour un individu donné à prélever indifféremment des repas sanguins successifs chez plusieurs hôtes différents pour la réalisation des cycles gonotrophiques*. Là aussi, les mœurs microprédatrices s’accompagnent d’une capacité de dissémination augmentée, de colonisation d’aires géographiques nouvelles, mais aussi d’hôtes différents.
Certains microprédateurs hématophages sont aptères et sont, de ce fait, beaucoup moins mobiles. Ainsi la femelle adulte de la punaise de lit ponctionne-t-elle aussi le sang de différents individus, mais, non ailée, elle demeure confinée dans un périmètre plus restreint. Son spectre d’hôtes demeure cependant assez large (Cimex lectularius homme, rongeurs, chiroptères, oiseaux). Or on sait qu’elle peut être transportée entre autres par les personnes avec leurs effets personnels (vêtements, valises, …) (Reinhardt & Siva-Jothy 2007). Les tiques dures semblent représenter un cas particulier, s’attachant solidement à l’hôte pour la réalisation d’un seul long repas, mais se laissant tomber pour pondre ensuite dans l’environnement. La femelle adulte de la plupart des tiques dures ne fait qu’un seul repas de sang, mais l’ampleur du prélèvement semble contrebalancer l’absence d’itération (Prasad 1987). Le repas, d’une durée moyenne de 43 jours chez Ixodes scapularis par exemple (Troughton et Levin 2007) induit une extension du corps de l’acarien de plus de 100 fois son volume initial et permet l’accomplissement de sa ponte unique. Un repas interrompu peut très difficilement être continué sur un autre individu hôte, même de la même espèce. D’une manière générale, un repas de sang interrompu quel que soit le stade chez les tiques est difficilement repris et mène souvent à la mort (ex. chez Dermacentor variabilis ; Amin et Sonenshine 1969). Le caractère unique du repas de sang chez la femelle adulte rend la distinction microprédateur/ectoparasite malaisée suivant la définition de Kuris et Lafferty (2000). Toutefois, son implication dans l’environnement la rapproche des microprédateurs. Quant à la plupart des puces (Siphonaptera : Ctenocephalidae, …), aptères par nature, et bien que le stade adulte demeure souvent longtemps sur l’hôte, pouvant aller jusqu’à se fixer solidement (Tungidae), elles présentent une biologie relativement proche de celle des moustiques. En effet, les larves ne sont pas hématophages (saprophages) et le développement complet de l’individu se déroule dans l’environnement.
c - Difficultés dans l’appréhension des associations lâches Les couples hôte-parasite typique, hôte-endosymbiote, par l’étroitesse de leur association, impliquent par excellence des contraintes adaptatives telles qu’ils représentent souvent des cas d’école et que des règles très strictes peuvent en être tirées (ex. le cas des gauphres à poches du genre Geomys et de leurs poux mallophages ; Page et Hafner 1996). La spécificité d’hôte chez ces organismes est par conséquent très élevée en général, intégrant de fréquents événements de coévolution dans l’histoire de leur association. A tel point, d’ailleurs, que certains auteurs en viennent à considérer la phylogénie des uns comme patron pour reconstruire celle de autres (ex. Verneau et al. 2002). Mais, dans les cas d’associations plus lâches, telle celle du microprédateur avec ses macroproies, au spectre d’hôtes par excellence élargi, et dans lesquelles le rôle de l’environnement extérieur à l’hôte est important, la régularité des processus adaptatifs et des patrons de variabilité a des chances d’apparaître bouleversée. Non seulement la complexité des interactions au sein de systèmes de ce type rend leur investigation déconcertante, mais encore les outils développés pour les reconstructions cophylogénétiques sont d’utilisation délicate dans le cadre de systèmes lâches. La réconciliation d’arbres de parasites associés à des hôtes multiples demeure par excellence malaisée. C’est d’ailleurs sans doute ce qui prévaut au conflit Dowling vs Page et Charleston (Dowling 2002, Page et Charleston 2002, Brooks et al. 2004) quant aux modèles les plus appropriés pour l’exploration de la cophylogenèse entre un parasite et son hôte. Le premier, 17 spécialiste des Dermanyssoidea, vaste groupe d’acariens principalement prédateurs, dont plusieurs lignées indépendantes semblent avoir évolué vers le parasitisme hématophage, de type principalement microprédateur, malgré quelques exceptions (Dowling 2006a, b), ne peut s’accommoder des mêmes outils que les seconds, spécialistes d’insectes parasites typiques (poux mallophages). Pour Page et Charleston, la réconciliation entre l’arbre des hôtes et l’arbre des parasites se doit de favoriser le plus grand nombre d’événements de coévolution. Pour Dowling, le critère du maximum de coévolution tendant à induire beaucoup de bruit dans une analyse intégrant des acariens aux mœurs plutôt prédatrices - et par conséquent plus opportunistes que les ectoparasites typiques - pose problème. Les événements les plus récurrents sont en effet probablement les transferts d’un hôte/proie à un autre, événements très difficiles à gérer s’ils ne se limitent pas à apparaître occasionnellement. Banks et Paterton (2005), explorant les difficultés d’investigation des systèmes hôte-parasite en cas d’hôtes multiples, remarquent que la plupart des études de cophylogenèse entre taxon hôte et taxon parasite traitent de parasites très spécifiques, arguant de la difficulté des reconstructions cophylogénétiques entre hôte et parasite à hôtes multiples.
d - Intérêt de l’étude de systèmes microprédateur aptère / macroproie L’étude d’associations entre consommateurs au spectre large et leur cible est pourtant d’un grand intérêt, puisqu’elle permet l’appréhension de processus adaptatifs complexes entre paramètres biotiques et abiotiques extrêmement divers. Le degré de mobilité intrinsèque du microprédateur vient contraindre son attachement à l’environnement de l’hôte. Les systèmes microprédateur aptères / macroproies fournissent un matériel d’étude particulièrement intéressant, car intermédiaires entre systèmes hôte / ectoparasites typiques et systèmes prédateurs / proies : l’impact de l’environnement est probablement plus important que chez l’ectoparasite typique, mais des contraintes locales (microenvironnementales) peuvent jouer un rôle et contraindre l’association si l’on compare avec les microprédateurs ailés.
La nature isolée des populations de parasites typiques fait que la dispersion est, si ce n’est très réduite, au moins étroitement corrélée à celle de l’hôte (ex. Blouin et al. 1995). La structuration spatiale des populations de microprédateurs, ou tout au moins des arthropodes hématophages accomplissant leur développement dans l’environnement, est potentiellement très différente de celle des parasites typiques. Et elle est nécessairement très différente entre un arthropode ailé et un arthropode aptère. La capacité intrinsèque de dispersion d’un microprédateur ailé peut être très importante. Gorrochotegui-Escalante et al. (2000) montrent par exemple un isolement à partir de distances de 90-250 kilomètres chez le moustique Aedes aegypti. Les microprédateurs aptères tels les punaises de lit s’éloignent de l’hôte à la différence des ectoparasites typiques (séjournant à quelques mètres de distance de l’hôte), mais leur capacité de dispersion intrinsèque est nécessairement très réduite comparée aux insectes ailés. Ce sont précisément ces deux compétences combinées qui permettent à la police scientifique d’utiliser le sang contenu dans les punaises de lit comme indice dans certaines enquêtes judiciaires (Szalanski et al. 2006).
Les populations d’ectoparasites typiques tendent vers une structuration spatiale concordant avec celle de l’hôte, alors que celle des microprédateurs ailés est marquée par des caractéristiques macroenvironnementales (climat, paysage) (ex : Paupy et al. 2005). Qu’en est-il de microprédateurs à mobilité plus réduite ? Inaptes à couvrir des distances importantes par leurs propres moyens, et a priori moins enclins à voyager sur l’hôte, ils pourraient être paradoxalement les moins mobiles des trois catégories. Ils présentent d’ailleurs souvent une capacité de résistance au jeûne élevée, leur permettant de survivre dans un environnement déserté et d’attendre le retour 18 ou la venue d’une macroproie (jusqu’à 9 mois chez D. gallinae, selon Nordenfors et al. 1999, six à douze mois chez Cimex lectularius selon Koehler et al. 2008). Les puces et les tiques dures, dont la station sur l’hôte est relativement élevée constituent des modèles intermédiaires entre ectoparasites typiques et microprédateurs aptères. En effet, les jours voire semaines nécessaires au gorgement complet des femelles adultes dans le genre Ixodes, la station sur l’hôte des puces adultes du genre Ctenocephalides par exemple augmente sensiblement la probabilité du transport par l’hôte au cours d’un de ses déplacements. Ces espèces sont aussi douées de capacité importante de résistance au jeûne, à différents stades de leur développement. Les systèmes microprédateur aptère – macroproie représentent des modèles fondamentalement différents des systèmes ectoparasite typique / hôte et des systèmes microprédateur ailé / macroproie. Leur implication dans le microécosystème de l’environnement de l’hôte en fait des modèles de grand intérêt pour l’exploration des interactions au sein d’une association a priori relativement lâche entre les deux organismes, mais potentiellement étroite entre le microprédateur et le microécosytème de l’environnement de la macroproie. Combes (2000) définit deux filtres génétiques primordiaux. Ces deux filtres représentent les caractéristiques impliquées (1) dans la rencontre entre parasite et hôte potentiel, (2) dans la compatibilité post-rencontre (nécessaire à la durabilité du système). L’étude de processus liés aux relations de descendance au sein d’un taxon microprédateur aptère et des contraintes évolutives afférentes au microécosystème du nid laisse envisager une meilleure compréhension de phénomènes adaptatifs complexes, à une échelle dépassant celle du seul hôte. Les deux filtres de Combes sont, chez les microprédateurs aptères, en prise avec des paramètres plus distants de l’hôte que chez l’ectoparasite typique. Complémentaire d’études portant sur des ectoparasites plus typiques, elle pourrait apporter de précieuses indications quant au rôle de paramètre abiotiques, ou à celui de la composition des communautés d’organismes très apparentés au parasite (arthropodes). C’est ainsi, par exemple, que Cimex lectularius, la punaise de lit la plus connue, semble avoir échappé à un goulot d’étranglement génétique, qui caractérise pourtant des parasites typiques, plus exposés aux pesticides dans les années 1940-50 aux Etats-Unis, selon Szalanski et al. (2008).
1.3 Modèles microprédateur aptère / oiseau A l’instar des punaises de lit, les femelles adultes de plusieurs espèces du genre Dermanyssus (Acari : Mesostigmata) sont microprédatrices. Principalement inféodés aux oiseaux, ces acariens constituent un intéressant modèle d’association microprédateur/macroproie. En effet, composante importante de l’arthropodofaune des nids d’oiseaux, ils sont souvent recensés dans ces îlots aux communautés d’arthropodes souvent riches et variables en fonction de l’espèce d’oiseau (Zeman and Jurík 1981, Burtt et al. 1991, Fenća and Schniererová 2004, Nosek and Lichard 1962, Fain and Galloway 1993, Majka et al. 2006, Merkl et al. 2004…). En outre, le caractère ailé de l’hôte, ses habitudes migratrices éventuelles pourraient contribuer à la dissémination plus large du microprédateur aptère, en cas d’utilisation de l’hôte comme véhicule. A l’heure actuelle, peu d’études ont porté sur la spécificité d’hôte et les voies de dissémination des populations de microprédateurs aptères inféodés à des hôtes ailés. La principale est celle menée par McCoy et coll. depuis plusieurs années sur les populations d’I. uriae, tique dure parasitant des oiseaux. McCoy (2001) et McCoy et al. (2003) ont montré l’impact de la biologie de l’hôte sur la différenciation entre populations chez cette tique d’oiseaux. Le modèle de cette étude étant une tique dure, c’est-à-dire aux habitudes hématophages strictes, quel que soit le stade, comme les espèces du genre Dermanyssus, mais dont la durée de la station sur l’hôte est nettement accrue, il est possible que l’impact des mouvements de l’hôte soit plus important que chez la plupart des dermanysses. Par ailleurs, l’environnement strictement naturel des hôtes de ces
19 tiques dans l’étude sus-citée (oiseaux marins, nichant dans des aires strictement naturelles) ne permet pas d’incriminer quelque action humaine que ce soit. Cela permet une analyse de la structuration des populations d’un microprédateur aptère inféodé à un hôte ailé hors de toute action humaine. L’omniprésence du genre Dermanyssus dans l’avifaune sauvage, ainsi que l’existence au sein des espèces du genre Dermanyssus d’au moins une espèce d’importance économique en font un modèle complémentaire fort intéressant. Il permet en effet l’investigation comparative de populations de microprédateurs aptères inféodés à des hôtes ailés en milieu sauvage et anthropisé. Cela laisse espérer non seulement des éclairages quant aux contraintes présidant à l’évolution d’associations relativement lâches, mais encore une meilleure compréhension de l’impact de l’activité humaine sur des associations plus ou moins liées à l’homme.
20
2 Introduction La classification des oiseaux suivie ici est celle de Peterson (2007). 2.1 Contexte : le genre Dermanyssus et le groupe gallinae Le genre Dermanyssus Dugès, 1834 (Acari: Mesostigmata: Dermanyssoidea: Dermanyssidae – cf. Annexe 1) regroupe des espèces hématophages ectoparasites d'oiseaux. Il est au début de la présente étude composé de 23 espèces décrites et a été divisé en deux sous-genres renfermant trois groupes par Moss (1968, 1978 ; cf. Fig. 1) : le sous-genre Dermanyssus scindé en 2 groupes (groupe gallinae -14 espèces - et groupe hirsutus – 4 espèces) et le sous-genre Microdermanyssus (5 espèces).
Genre Sous-genre Espèces groupe gallinae D. antillarum Dusbabek and Cerny, 1971 D. chelidonis Oudemans, 1939 D. faralloni Nelson and Furman, 1967 D. gallinae De Geer, 1778 D. gallinoides Moss 1966 Dermanyssus D. hirundinis (Hermann, 1804) D. prognephilus Ewing, 1933 Moss, 1967 D. transvaalensis Evans and Till, 1962 D. triscutatus Krantz, 1959 D. trochilinis Moss, 1978
groupe hirsutus D. grochovskae Zemskaya 1961 Dermanyssus Dugès, 1834 D. hirsutus Moss and Radovsky 1967 D. quintus Vitzthum, 1921
Microdermanyssus Moss, 1967 D. alaudae (Schrank, 1781) D. americanus Ewing 1922 D. brevis Ewing, 1936
Incertae sedis D. longipes Berlese and Trouessart, 1889 (incertae sedis) D. passerinus Berlese and Trouessart, 1889 (incertae sedis)
Figure 1. Aperçu de la classification et de la composition du genre Dermanyssus au début de l’étude, selon Moss (1978).
Les espèces du groupe gallinae sont douées d’une réelle capacité de gorgement – avec extension importante des organes digestifs et de l’opisthosome*1- qui leur permet de prélever en quelques minutes un repas de sang suffisant pour accomplir une métamorphose ou une ponte, selon le stade concerné (Radovsky 1994). Leur cycle de vie comporte cinq stades (cf. Fig. 2) : oeuf, larve, protonymphe, deutonymphe et adulte. Seuls les trois derniers ont besoin de se nourrir de sang, les protonymphes et deutonymphes pour accomplir leur métamorphose (un repas chacune
1 La définition des mots suivis d’un astérisque (*) est consultable dans le lexique, en fin de document (p. 255). 21 seulement), les femelles adultes avant chaque ponte pour la maturation de leurs œufs (cycles gonotrophiques*). La biologie des autres espèces (groupe hirsutus du sous-genre Dermanyssus et sous-genre Microdermanyssus) a été peu étudiée. Toutefois, plusieurs éléments de la littérature laissent penser que les espèces du groupe hirsutus, dont la morphologie est plus adaptée à l’agrippement qu’à la course (pattes massives et courtes), tout au moins certaines d’entre elles, ont des habitudes plus caractéristiques des parasites typiques, tels les poux, par exemple, qui passent leur vie sur l’hôte. Ainsi, D. grochovskae et D. quintus demeurent sur l’hôte, pondent directement dans ses plumes et ne présentent pas à proprement parler de capacité de gorgement. Leurs repas sont de faible ampleur et répétés au cours d’un même stade nymphal (Moss 1978). Les espèces du sous-genre Microdermanyssus présentent peut-être une biologie intermédiaire, nidicole durant la période de nidification de l’hôte, stationnant sur l’hôte durant les périodes hivernales : Zemaskaya (1968) et Zemskaya et Ilienko (1958) signalent la présence en nombre bien plus important d’individus appartenant à D. americanus, accompagnés d’œufs, sur l’hôte en hiver qu’au cours de la nidification. Zemskaya (1971) estime, sur la base d’observations portant sur deux espèces du groupe hirsutus, deux du groupe gallinae et une du sous-genre Microdermanyssus qu’une transition depuis le mode de vie nidicole vers le mode parasite permanent (parasite typique) est manifeste dans le genre Dermanyssus. Toutefois ce qui permet d’orienter le sens d’évolution (du groupe gallinae vers les autres groupes) n'apparaît pas clairement, aucune analyse phylogénétique ne venant étayer ce sens. En bref, au sein du genre Dermanyssus, les espèces du groupe gallinae ont un mode de vie relativement déconcertant si l’on compare à celui de la majorité des parasites connus et aux deux autres groupes du même genre. A vrai dire, si l’on considère les femelles adultes au moins dans le groupe gallinae, elles se comportent davantage en microprédateurs qu’en parasites typiques : ponctionnant du sang avant chaque ponte à plusieurs reprises au cours de leur vie, elles ne se nourrissent pas nécessairement sur le même individu, à l'instar des femelles adultes de moustiques ou des punaises des lits. Cela correspond aux caractéristiques des microprédateurs, partagées par les moustiques, les punaises hématophages, et d’autres animaux zoologiquement très différents comme les sangsues, si l’on se réfère aux catégories des consommateurs de Kuris et Lafferty (2000).
22 Figure 2. Cycle de vie des espèces du groupe gallinae.
a - Une espèce d’importance économique dans un genre méconnu D. gallinae, improprement nommé communément pou rouge des poules, est connu pour ses dégâts majeurs en élevage de poules pondeuses, entraînant d’importantes pertes économiques. Cette espèce sévit dans les élevages principalement en Europe et induit problèmes sanitaires et pertes financières (cf. plus-bas § 2.1b - ). Sa prévalence est forte en Europe : environ 80% des élevages sont infestés en France. Elle est en outre croissante en Amérique du Sud (Tucci et al. 2008). Les élevages d'Amérique du Nord semblent en être exempts, ou presque, pour l'heure (B. Mullens, comm. pers., Phillis et al. 1976). Dans cette région du monde, une espèce appartenant à une famille apparentée semble occuper cette niche : Ornithonyssus sylviarum (Canestrini and Fanzago, 1877) (Dermanyssoidea: Macronyssidae) (Axtell and Arends 1990, Mullens et al. 2001). Mais le mode de vie de cette espèce est plus proche de l’ectoparasite typique que du microprédateur, car, tout au moins chez les volailles d’élevages (Mullens et al. 2001) et chez les canaris (observation personnelle), les individus passent le plus clair de leur temps sur l’hôte et pondent directement dans le plumage. Cela suggère qu’en fait les deux espèces d’importance économique occupent deux niches chevauchantes, si ce n’est différentes.
D’une manière générale, peu d’études ont porté sur le genre Dermanyssus, tant sur le plan taxinomique qu’écologique. Or les déprédations non négligeables causées par au moins l'une des espèces qui le composent, ainsi que les difficultés de traitement associées suscitent aujourd'hui une certaine mobilisation dans les institutions et universités touchant aux sciences vétérinaires en Europe. Ainsi, de nombreux essais à visée appliquée sur l'espèce D. gallinae sont publiés régulièrement et connaissent un essor notoire depuis 2005 : 1-3 publications par an avec quelques rares pics à 4, 5 ou 6 jusqu'à 2004, puis 6 en 2005, 6 en 2006, 12 en 2007, 16 en 2008, 21 en 2009 (PubMed (NCBI), mot-clé "Dermanyssus"). Ces publications portent pour la plupart sur des essais de sensibilité à diverses molécules de synthèse et huiles essentielles (ex: Beugnet et al. 1997, Todisco et al. 2008). Certaines traitent aussi de contraintes environnementales liées au cycle de vie
23 (Nordenfors et al. 1999, Tucci et al. 2008), de travaux immunologiques à visée vaccinale (Nisbet et al. 2006, Wright et al. 2009), ainsi que du rôle vecteur vis-à-vis de pathogènes (Valiente Moro et al. 2005, 2007). Enfin, une équipe danoise travaille sur les stimuli présidant aux déplacements de l’acarien (Kilpinen 2001, 2005).
Mais aucune étude n'a clairement défini préalablement l'espèce cible. Malgré les dégâts engendrés en élevage avicole par D. gallinae, la classification du genre Dermanyssus au niveau spécifique est restée très confuse. Alors que sa description date de 1833, ce n’est qu’à partir des années 1960 que quelques auteurs ont commencé à réviser le genre. Une discussion constructive entre Krantz et Sheals sous forme d’articles (1959-1962) commença à clarifier la définition du genre, tout au moins vis-à-vis des autres genres apparentés. Evans et Till publièrent en 1962 la première révision complète du genre au niveau spécifique. Moss commença en 1967 un travail d’investigation des relations entre les espèces au sein du genre mettant à profit une approche phénétique. Son travail apporta de précieux éléments, avec des subdivisions internes (sous-genres Microdermanyssus et Dermanyssus, ce dernier incluant les groupes gallinae et hirsutus), mais demeura inachevé, sa dernière publication (Moss 1978) concluant davantage sur des expectatives (une étude annoncée comme en cours) que des clarifications. Moss (1978) insista aussi beaucoup sur l’extrême variabilité des caractères morphologiques au sein d’une même population et mit vivement en garde les utilisateurs de sa clé dichotomique contre les risques importants d’erreur. Enfin, avant la présente étude, le genre n’avait fait l’objet d’aucune reconstruction phylogénétique prenant en compte des caractères moléculaires. La délimitation interspécifique demeure donc peu claire, en particulier dans le groupe gallinae sensu Moss (1978).
b - Impact direct sur l’hôte L’impact de D. gallinae sur son hôte dans les élevages de volaille, est relativement important. Il induit le déclassement des œufs tachés par les acariens écrasés (cf. Annexe 2a) et est potentiellement capable de transmettre des agents pathogènes : protozoaires (Lainson 1960), bactéries et/ou virus pathogènes (Valiente Moro et al. 2005, 2007, 2009). En outre, perturbant le sommeil des poules, il génère du stress, qui se traduit entre autres par une baisse du rendement (augmentation de la consommation d’aliment non accompagnée d’une augmentation de la production), une détérioration du plumage par augmentation du lissage des plumes (Kilpinen 1999, Kilpinen et al. 2005). Lors d’infestations massives, une chute de la ponte, une perte de poids et une augmentation de la mortalité peuvent apparaître rapidement (Kilpinen et al. 2005). A plus long terme, une modification des valeurs de certains paramètres sanguins a pu être notée dans certains cas, témoignant probablement d’une anémie régénérative (Kirkwood 1967, Keçeci et al. 2004). Au laboratoire, nous avons pu constater à plusieurs reprises la mort en une nuit par exsanguination de jeunes poulets (âgés de 3 à 15 jours) placés au contact d’aggrégats importants de D. gallinae. Dans la faune sauvage, Clayton et Tompkins (1994, 1995) ont montré qu'une prolifération de D. gallinae pouvait induire une réduction notoire du temps de couvaison des œufs et avoir un effet délétère important sur le succès de la reproduction du pigeon biset (Columbia livia: Columbidae). Seules deux autres espèces de Dermanyssus ont fait l'objet d'études traitant de leur impact sur l’hôte : Moss et Camin (1970) ont démontré que les nids d'hirondelle noire (Progne subis: Hirundinidae) parasités par D. prognephilus produisaient des poussins moins lourds que les nids non parasités. En revanche, l’impact de D. hirundinis sur le troglodyte familier (Troglodytes aedon: Certhiidae) n’a pas encore pu être clairement mis en évidence (Johnson et Albrecht, 1993; Pacejkaa et al. 1996, 1998). Il semblerait que l’impact de D. hirundinis, tout au moins chez le trogolodyte familier, soit moins flagrant que celui de D. gallinae chez la poule et chez le pigeon biset ou que celui de D. prognephilus chez l’hirondelle noire. 24 c - Distribution
2.1.c.1 Spectre d’hôte La spécificité d’hôte des espèces du genre Dermanyssus était, au début du présent travail, réputée très faible dans la majorité des espèces. Notamment, plus de 40 espèces d’oiseaux réparties dans 8 ordres différents ont pu être recensées comme hôtes pour la seule espèce D. hirundinis (Hermann, 1804) et plus de 30 espèces d’oiseaux réparties dans 6 ordres différents pour D. gallinae (De Geer, 1778). D. prognephilus Ewing, 1933 était répertorié chez 2 familles de Passériformes et 1 de Piciformes. Seules certaines espèces telles D. quintus Vitzthum, 1921, D. hirsutus Moss & Radovsky, 1967, D. alaudae (Schrank, 1781), étaient connues pour ne parasiter qu’une seule famille d’oiseau (respectivement Picidae, Picidae, Alaudidae). Pour d’autres, rencontrées trop rarement jusqu’à présent, l’ampleur du spectre d’hôtes est très difficile à évaluer (D. wutaiensis Gu et Ting, 1992 et D. brevirivulus Gu et Ting, 1992, de D. grochovskae Zemskaya, 1961, D. antillarum Dusbabek & Cerny, 1971, D. trochilinis Moss, 1978, D. rwandae Fain, 1993, D. nipponensis Uchikawa et Kitaoka, 1981, …).
2.1.c.2 Répartition géographique Outre une large répartition dans la taxinomie des oiseaux, plusieurs espèces du genre Dermanyssus, aussi bien dans le groupe hirsutus que dans le groupe gallinae présentent aussi une large répartition géographique. D. quintus est recensé en Europe (Vitzthum, 1921), en Russie (Zemskaya 1971), en Amérique du Nord (Moss et al. 1970). D. hirundinis est abondamment signalée dans différents pays d’Europe (Zeman et Jurík 1981, Fenća et Schniererová 2004, 2005, Evans et Till 1966, …) et en Amérique du Nord (Moss et al. 1970, …). D. gallinae, seule espèce du genre Dermanyssus connue pour parasiter les volailles domestiques, est cosmopolite, recensé aussi bien dans le Nouveau Monde que dans l’Ancien monde, aussi bien dans la faune sauvage (FS) qu’en élevage (E) : Amérique du Nord : Moss et al. 1970 (FS), Amérique du Sud : Tucci et al. 2008 (E), Europe : Rép. Tchèque, Zeman et Jurík 1981 (FS), Slovaquie, Fenća et Schniererová 2004, 2005 (FS), Royaume Uni, Guy et al. 2004 (E), Royaume Uni, Italie, Pays-Bas, …Sparagano et al. 2009 (E), Afrique (Maroc, Sahibi et Rhalem 2007; Egypte, El Kady et al. 1995 (E)), Asie (Israel, Rosen et al. 2002 (E, FS), Turquie, Kececi et al. 2004 (E)), Chine (Gu et Ting 1992), Japon (Uchikawa et Takahashi 1985 (FS)). Certaines espèces sont notées dans une seule localité/région, mais il s’agit dans la plupart des cas d’espèces rencontrées ponctuellement (1-3 collectes en tout, dans la même région, à la même période). C’est le cas de D. wutaiensis et D. brevirivulus (Chine), D. antillarum (Cuba), D. trochilinis (Pérou), D. rwandae (Rwanda), D. nipponensis (Japon). Dans la plupart de ces cas, l’endémisme est peu probable, l’aire de répartition demeurant obscure par manque de données. D. grochovskae, décrit en Russie a aussi été rencontré au Japon (Uchikawa et Takahashi 1985). D. carpathicus Zeman, 1979, jusqu’alors noté uniquement en République Tchèque, le pays type, a d’ailleurs été isolée à de nombreuses reprises dans des nids collectés en France dans la présente étude. D. americanus Ewing, 1923 est peut-être réellement inféodé à l’Amérique du Nord (Ewing 1923, Phillis 1972).
d - Reproduction Le mode de reproduction est dans le genre Dermanyssus difficile à explorer, du fait (1°) de son hématophagie stricte avec nécessité de piquer à travers une membrane et de stimuli alimentaires encore mal maîtrisés et (2°) de la faible différentiation morphologique entre les stades / sexes hors préparation microscopique, c’est-à-dire sur individus vivants. 1°) Les difficultés de la nutrition in vitro rendent les explorations individuelles très difficiles, un individu pouvant très difficilement être amené au stade adulte avec conservation
25 assurée de la virginité, afin de tester la parthénogenèse et de réaliser des accouplements contrôlés. Pour l’heure, les essais de gorgement individuels in vitro comme in vivo n’ont pas encore permis le développement d’un individu isolé jusqu’au stade adulte et ne permettent actuellement que le développement de groupes d’individus (ex : Valiente Moro 2007, McDevitt et al. 2006, expérience personnelle). 2°) En outre, le sexage d’individus vivants est très difficile, même à la loupe binoculaire, les caractères sexuels n’apparaissant clairement qu’après préparation microscopique (orifice spermatique et spermadactyle chez le mâle adulte, rabat de l’ovipore chez la femelle adulte, notamment). Un éclairage très rasant à un fort grossissement à la loupe binoculaire peut permettre la séparation de stades et sexes par observation des plaques ventrales et dorsales chez les plus grosses espèces, comme D. gallinae, mais la vivacité de ces acariens rend l’opération très délicate sur individus vivants. En outre, si cette méthode permet sans trop d’échec la distinction entre mâles et femelles (plaques ventrales et dorsale beaucoup plus développées chez le mâle ; fig. 3), la discrimination entre femelles adultes et deutonymphes demeure très délicate. En effet, les plaques dorsales étant identiques, il faut réussir à percevoir le rabat de l’ovipore propre à la femelle adulte (plaque ventrale entière chez la deutonymphe, scindée en la très mince plaque sternale à l’avant et la plaque épigyniale à l’arrière, séparées par le rabat fripé de l’ovipore chez la femelle adulte ; fig. 3). C’est pourquoi peu d’études expérimentales ont été menées sur cet aspect.
A. B.
26 C.
Figure 3. Caractères permettant la séparation des stades/sexes chez D. gallinae. A. Caractères sexuels primaires et secondaires d’un mâle adulte visibles en microscopie photonique. La flèche blanche désigne l’orifice spermatique du mâle. Les contours de la plaque anale, surlignée en jaune, peuvent être aperçus à la loupe binoculaire, avec un éclaraige rasant. B. Caractère sexuel secondaire d’une femelle adulte. La flèche blanche indique l’emplacement de l’ovipore (cf. D). Surlignage en jaune : plaque anale, cf. ci-dessus. C. Caractère sexuel primaire d’une femelle adulte vu en microscopie photonique (à gauche) et électronique à balayage (à droite) : rabat de l’ovipore.
Les principaux auteurs qui ont étudié la reproduction chez Dermanyssus (et apparentés) sont Oliver et Hutcheson (Oliver 1966, 1977, Hutcheson et Oliver 1988). Ils ont développé un sexage relativement efficace par observation au « microscope à dissection » des plaques ventrales sur des individus maintenus dans des tubes en verre (difficulté n°1 ci-dessus). Ils ne décrivent pas avec beaucoup de précision ce qu’ils observent2, mais on peut penser que pour le discernement entre deutonymphe et femelle adulte, outre l’observation des plaques, ils retenaient simplement le critère taille (femelles = les plus grosses). Ils ont en outre partiellement résolu la difficulté n° 2 ci- dessus en partant du principe que les deutonymphes séparées du reste de la population testée juste après le repas préalable à la mue imaginale n’avaient pas pu être fécondées. Ils procédaient au nourrissage des acariens en groupe sur jeunes poulets. En bref, il ressort des études d’Oliver et Hutcheson que chez D. gallinae (1) les œufs qui donneront naissance à des mâles sont non fécondés et haploïdes, (2) que les œufs qui donneront naissance à des femelles sont fécondés et diploïdes, (3) que l’accouplement (pas la fécondation) et le gorgement (jusqu’à un point minimum critique) sont des préalables nécessaires à l’oviposition, (4) qu’un seul accouplement suffit pour
2 Par exemple, selon Hutcheson & Oliver (1988) « The sex of unfed adults was determined detecting presence or absence of the female genital plate as viewed through a 19 by 48-mm glass vial with a dissecting microscope ». On peut supposer que la “plaque génitale” observée est en fait le rabat membraneux de l’ovipore, la plaque épigyniale étant similaire à la plaque ventrale des deutonymphes. Mais il s’agit sans doute en fait des individus les plus gros, présélectionnés sur leur taille, sur lesquels il recherchait la plaque épigyniale de la femelle, par opposition à la plaque holoventrale du mâle. Toutefois, les mâles, naturellement plus petits que les femelles, peuvent ressembler aux deutonymphes. 27 féconder les œufs femelles de toutes les pontes d’une seule femelle, (5) mais que les œufs fécondés le sont rapidement après l’accouplement, et demeurent stockés tels quels. Ainsi, apparemment, les spermatozoïdes eux-mêmes ne sont pas stockés comme chez de nombreux arthropodes équipés de spermathèques (ex. cigales), ceux qui n’ont pas fécondé d’œuf disparaissant en quelques jours (>50j in Oliver 1966, <2j in Hutcheson et Oliver 1988). Dans le genre Dermanyssus, Oliver et coll ont réalisé des expérimentations sur la reproduction chez D. gallinae seulement. Ils ont toutefois aussi vérifié l’haplodiploïdie chez D. prognephilus. Il apparaît donc, selon Oliver et Hutcheson, que D. gallinae est arrhénotoque*. Même si l’accouplement est nécessaire à l’induction de l’oviposition, il s’agit là de pseudogamie, les œufs mâles demeurant non fécondés. La question de la pseudoarrhénotoquie* (nommée aussi para-haploïdie), toutefois, a été soulevée par Dotson (1982). Cette petite étude menée dans le cadre d’un mémoire de master n’a pas permis de mettre en évidence de pseudoarrhénotoquie. Elle a par ailleurs fait apparaître une augmentation constante du nombre d’œufs au fil des cycles gonotrophiques*. Oliver (1966) avait noté une augmentation jusqu’au 6ème cycle, mais une décroissance aux 7ème et 8ème. Dotson (1982) relève aussi une évolution du sexe ratio au fil des pontes (premières pontes seulement mâles, dernières seulement femelles, intermédiaires mixtes), mais cela n’a pas été vérifié ni confirmé par une publication. Quant à la supposée pseudoarrhénotoquie, l’inférence phylogénétique de Cruickschank et Thomas (1999) sur la base de l’ADNr 28S, visant à appréhender l’évolution de la ploïdie au sein des Dermanyssina sur la base de 10 taxa spécifiques tend à confirmer aussi le statut arrhénotoque expérimentalement déjà mis en évidence par Oliver et Hutcheson et non invalidé par Dotson (1982). Le sexe chez D. gallinae semble donc simplement déterminé par la fécondation ou non de l’œuf, comme chez de nombreux autres arthropodes, tels les abeilles. Cela est en outre probablement le cas chez les autres espèces du genre, puisque les autres espèces testées de diverses familles de Dermanyssina se sont avérées arrhénotoques aussi (Oliver 1966, 1977).
Enfin, la fécondité individuelle de l’acarien, si l’on compare à d’autres acariens (voire à d’autres arthropodes) déprédateurs, est relativement faible. Une femelle adulte peut réaliser jusqu’à 8 cycles gonotrophiques* (Wood 1917) au cours de sa vie, durant lesquels elle peut pondre 1 à 8 œufs, pour un total moyen de 23 œufs (Oliver 1966) vs plusieurs milliers à plusieurs dizaines de milliers d’œufs par femelle chez certaines tiques (une seule ponte par femelle), 82 à 439 œufs par femelle chez Cimex rotundatus (Dunn 1924) ou 60 à 200 oeufs par ponte chez les moustiques (Anderson & Harrington 2009).
e - Particularités de la biologie des espèces du genre Dermanyssus en lien avec les difficultés de traitement rencontrées en élevage
2.1.e.1 Variabilité du temps de génération Malgré sa faible fécondité, D. gallinae parvient à envahir rapidement des élevages de manière plus ou moins délétère. La rapidité du cycle vient contrecarrer la taille réduite des pontes. En effet, en conditions optimales de température, hygrométrie et disponibilité d’hôte (généralement réalisées dans les élevages de volailles), l’acarien peut accomplir une génération complète d’œuf à œuf en une semaine. Nordenfors & Hoglund (2000) ont montré entre autres qu’une fois l’acarien détecté dans un élevage (méthode de piégeage standardisé), il lui fallait en moyenne 5 mois pour atteindre un niveau de population à l’équilibre, envahissant la totalité de l’élevage en Suède. Ce délai de plusieurs mois entre le début de l’infestation et le pic de
28 prolifération est l’une des raisons de la quasi-absence de problèmes engendrés par ledit pou rouge en élevages industriels de chair (bandes* de quelques semaines, entre 2 vides* sanitaires). Mais une fois l’équilibre atteint, et même un peu avant (à partir d’une infestation moyenne de 150000 à 200000 acariens par poule), l’impact sur les poules peut devenir important, allant jusqu’à perte de poids, anémie, voire augmentation significative de la mortalité (Kilpinen et al. 2005). Meyer- Kühling et al. (2007) ont aussi noté une augmentation de 400% en 42 jours d’une population de D. gallinae dans un bâtiment témoin (exempt de traitement) d’un élevage de pondeuses en Allemagne. Le temps de génération est donc rapide en conditions d’élevage. Entre deux bandes et dans la faune sauvage, toutefois, des pauses plus ou moins longues sont à signaler. En effet, en l’absence de tout hôte, D. gallinae peut survivre assez longtemps. Mais son développement complet (larve à adulte) requérant absolument 2 (chez le mâle) à 10 (chez la femelle réalisant le maximum de cycles gonotrophiques*) repas de sang par individu (selon le sexe, cf. Annexe 2a), la survie en condition de jeûne n’est marquée par aucune génération. Sa résistance au jeûne est importante et varie en fonction de la température et de l’hygrométrie relative. Des expérimentations de laboratoire ont mesuré cette capacité à survivre dans des conditions adverses (Wood 1917, Nordenfors et al. 1999). Protonymphes et deutonymphes, mâles adultes et femelles non encore gorgés peuvent survivre sans nourriture durant plusieurs mois (jusqu’à 9 mois pour certains individus selon Nordenfors 1999). La longévité des femelles adultes semble raccourcie une fois pris le premier repas de sang pour la maturation des œufs selon Wood (1917). Toutefois, les femelles qui se nourrissent et pondent sans interruption sembleraient vivre plus longtemps que celles qui, après un premier cycle gonotrophique*, se trouvent privées de nourriture et ne peuvent pas enchaîner un deuxième cycle. Dans la faune sauvage, Moss (1978) relève une capacité à survivre en l’absence d’hôte (l’hirondelle noire, Progne subis), durant la saison froide, jusqu’à 7 mois 1/2 chez D. prognephilus. Pacejka et Thompson (1996) ont mené des expérimentations de terrain et montré que D. hirundinis aussi est capable d’hiverner, dans les vieux nids de troglodyte familier de la saison précédente. Cela signifie que chez cette espèce aussi, certains individus survivent plus de 8 mois sans nourriture (et dans des conditions de températures très variables). Phillis (1972) montra en outre que les populations (indifférenciées) de D. hirundinis et D. americanus dans des nids de moineau domestique (Passer domesticus: Passeridae) étaient coordonnées avec le cycle court de l'oiseau (couvaison des œufs : 13-14 jours ; élevage des poussins : 17 jours selon http://www.oiseaux.net/) : les populations de ces deux espèces amplifieraient assez rapidement durant la période de pré-nichée, alors que le mâle construit le nid, ainsi que pendant la couvaison des œufs. Mais surtout, l’amplification deviendrait extrême durant le nourrissage des poussins, pour s’amortir rapidement et amorcer une décroissance rapide peu après le départ des oiseaux. Burtt et al. (1991), sur trois autres espèces de passériformes (l’hirondelle des arbre Tachycineta bicolor : Hirundinidae, le troglodytes familier Troglodytes aedon : Certhiidae, et le merlebleu de l’Est, Sialia sialis : Muscicapidae), mettent en évidence une dynamique comparable et soulignent l’extrême petitesse de la population fondatrice chez D. hirundinis. L’accélération de la croissance leur semble corrélée avec l’apparition des fourreaux des plumes chez les petits, dont ils supposent que les papilles, fortement irriguées, permettraient aux acariens de trouver des repas abondants et aisés. Une grande flexibilité dans le temps de génération, ainsi qu’une adaptation au cycle de l’hôte permettent ainsi aux espèces sauvages de développer des populations importantes, malgré la durée réduite de disponibilité de l’hôte au cours de l’année, comparée à celle des poules en condition d’élevage.
29 Aucune étude n’ayant porté précisément sur la biologie des espèces du groupe hirsutus et du sous-genre Microdermanyssus, nous n’avons pas d’information quant à leur résistance au jeûne. Toutefois, le mode de vie de D. grochovskae et D. quintus (cf. supra) ne laisse pas présager une nécessité de survie sur de longues périodes en l’absence d’hôte, puisqu’elles le suivent durant l’hiver (Zemskaya 1971). Il n’est pas improbable que la résistance au jeûne chez ces espèces qui ne se gorgent pas à proprement parler, mais réalisent de petits repas répétés et pondent sur l’hôte, soit réduite comparée à celles du groupe gallinae. La longévité d’O. sylviarum, aux habitudes apparemment similaires aux espèces du groupe hirsutus et du sous-genre Microdermanyssus, est nettement réduite comparée à celle de D. gallinae. En effet, les essais de Kirkwood (1963), comparant la longévité de ces deux parasites de volailles dans un poulailler vide, ont obtenu des individus de D. gallinae vivant après 8 mois, alors que chez O. sylviarum, les derniers vivants ont été notés durant la troisième semaine.
2.1.e.2 Etroite relation avec le microenvironnement (nid, litière) En outre, les habitudes particulières de D. gallinae, parasite aux mœurs de moustique aptère, ou plutôt de punaise des lits, en font un ennemi redoutable de l'éleveur. Ne demeurant que peu de temps sur l'hôte, le temps d'un rapide repas seulement, il se retranche rapidement dans des interstices étroits, nombreux dans les structures d’élevages. Dans la faune sauvage, les acariens du groupe gallinae sont très nettement plus nombreux dans les nids que sur l’hôte directement, à la différence du groupe hirsutus, où ils sont plus nombreux sur l’hôte en permanence chez D. hirsutus et D. grochovskae, et durant l’hiver chez D. americanus (Moss 1978, Zemskaya 1968). Les habitudes nidicoles des acariens du groupe gallinae sont doublées d’un thigmotactisme* certain, qui les pousse manifestement à s’agglutiner avec leurs congénères dans des espaces très étroits. Cette tendance est particulièrement visible lorsque l’on enferme un prélèvement de litière contenant des acariens du genre Dermanyssus dans un sachet hermétique en matière plastique transparente: après quelques heures, la majorité des acariens se trouve réunie au fond des angles du sachet (observation personnelle ; cf. Fig. 4). Ainsi, peut-être davantage stimulés par ce thigmotactisme que par leur célèbre lucifugie, ils s’amassent en pleine lumière dans la zone où les points de contacts avec le plastique sont les plus nombreux.
Figure 4. Agrégat de D. gallinae amassés au fond d’un angle de sachet plastique, en pleine lumière.
30 A l'instar de la punaise des lits Cimex lectularius dans les habitations humaines, l’acarien dans les élevages est par conséquent très difficile à atteindre avec les molécules acaricides existantes. Traiter l’oiseau est inutile, les traitements de l’environnement souvent insuffisants du fait de ses habitudes nidicoles avec propension à s’immiscer dans des espaces protégés. Et l'acarien, abrité, n'est en général détecté qu'une fois sa population largement développée au sein du bâtiment. La détection précoce des infestations à D. gallinae, comme celles à Cimex lectularius, est très difficile. A tel point que pour cette dernière, des méthodes développées aux Etats-Unis pour des établissements recevant du public utilisent des chiens détecteurs de punaise des lits (Koehler et al. 2008), ces punaises émettant des substances odorantes. Les chiens permettent ainsi de détecter les infestations débutantes, avant même que les clients ne s’en aperçoivent à leurs dépens. A ces difficultés d’ordre comportemental s’ajoutent d’importantes entraves d’ordre réglementaire, liées aux limites maximales de résidus (LMR), en particulier dans les œufs. Les LMR, déterminées par le règlement européen 2377/90/CEE3, rendent le traitement des élevages de pondeuses particulièrement délicat. Au début de cette étude, aucun traitement n’était autorisé durant la bande*, d’une durée d’un an dans cette filière. Seules quelques molécules (organophosphorés* et pyréthrinoïdes) étaient autorisées pendant le vide sanitaire*. C’est l’une des principales raisons pour lesquelles les pondeuses sont les volailles les plus affectées par cet acarien. En 2007, un produit à base de phoxime (organophosphoré*) a obtenu son autorisation de mise sur le marché. Il demeure aujourd’hui le seul produit composé de molécule de synthèse autorisé durant la bande en élevage de pondeuse standard en France. Dans un but alternatif, l’équipe d’I. Lesna (Université d’Amsterdam) travaille en collaboration avec la société Koppert Biological Systems à la mise en œuvre d’une démarche biologique impliquant un acarien auxiliaire (cf. Annexe 2b). Cette approche est aussi à l'étude dans l'équipe finlandaise de Tuovinen (2008). Les habitudes de chasse active à l’affût des espèce retenues (Mesostigmata : Laelapidae) en font des auxiliaires prometteurs, premier moyen de lutte susceptible d’agir sur les acariens agrégés dans les interstices abrités sus-décrits. Des cycles lumineux mimant des photopériodes très courtes (4h/2h par exemple) semblent aussi participer au contrôle de l’acarien (Sokóá et al. 2008). 2.2 Problématique Préalablement à toute étude d’ordre écologique portant sur un taxon spécifique, une caractérisation précise de ce qu’il représente - c’est-à-dire une délimitation entre les populations qui le composent et celles qui composent les taxa apparentés - constitue une étape indispensable. Dans le cas de D. gallinae, cet aspect demeure, au début de l’étude, complètement irrésolu. Par ailleurs, si certains éléments précis de la biologie de D. gallinae ont été appréhendés, son écologie et la dynamique de ses populations demeurent très obscures, en particulier les voies de dissémination de ses populations.
De nombreuses questions demeurent en suspens parmi les parties prenantes de la filière avicole. L'intensité et la ténacité de l’infestation dans certains élevages paraissent souvent inexpliquées. Souvent, d’autres élevages, dans des conditions apparemment similaires, maintiennent un niveau de population acceptable. Les cas d’élevages à double bâtiment mitoyen dans lesquels un seul bâtiment est infesté de manière récurrente, l’autre demeurant à peu près exempt de pou rouge (observation personnelle), aussi laisse perplexe. En outre, dans les cas d’élevage en plein air, les éleveurs et les vétérinaires se demandent souvent si les oiseaux sauvages
3 Un nouveau règlement sera applicable d’ici peu: le règlement (CE) no 470/2009 du Parlement européen et du Conseil du 6 mai 2009 abrogeant le règlement (CEE) no 2377/90. 31 ne sont pas la source de l’infestation (Eckert et al. 2005, Hoffmann 1987). En effet, certains passereaux (moineaux, mésanges, …) viennent parfois picorer aux côtés des poules. Une transmission des acariens par l’avifaune sauvage peut sembler improbable si l’on prend en considération les habitudes nidicoles du pou rouge. En revanche, le large spectre d’hôtes des espèces du groupe gallinae établi dans la littérature n’est pas sans renforcer cette suspicion. Enfin, le rôle des échanges commerciaux - en particulier l'introduction potentielle de l'acarien par les poulettes entrant dans un bâtiment pour former une nouvelle bande de pondeuses - restent méconnus.
Ces énigmes sont fortement compliquées par la faiblesse des informations taxinomiques fournies par la morphologie hautement variable au sein des populations dans le groupe qui englobe D. gallinae. Les seules espèces réputées spécifiques à une famille (cf. supra) appartiennent au groupe hirsutus (D. quintus, D. hirsutus) ou au sous-genre Microdermanyssus (D. alaudae), qui sont nettement plus caractérisées morphologiquement que celles du groupe gallinae. Evans et Till (1962) ont suggéré la présence de nombreuses erreurs dans les inventaires, la faible caractérisation morphologique poussant certains auteurs à assigner systématiquement le nom D. gallinae à tout individu lui ressemblant. Moss (1978) insiste sur les citations hautement suspectes dans la littérature du fait de l’occurrence de nombreuses espèces morphologiquement similaires.
Une spécificité d’hôte plus importante doit-elle être attendue d’espèces mal délimitées entre elles dans le groupe gallinae ? Avons-nous vraiment affaire à une seule espèce, D. gallinae, dans tous les élevages de pondeuses ? Et le groupe gallinae est-il composé d’un nombre d’espèces décrites qui surpasse le nombre des entités réellement isolées sur le plan reproducteur ? Ou bien, à l’inverse, les entités présentes en élevages appartiennent-elles à plusieurs espèces différentes, mal distinguées sur la base morphologique, seule disponible jusqu’alors ? Dans tous les cas, la réflexion et les stratégies de lutte contre ce ravageur aux importantes conséquences économiques sont fortement dépendantes de son identité précise.
En outre, aucune étude précise concernant les modes de dissémination des espèces du genre Dermanyssus n'est disponible au début du présent travail. Les observations de Valera et al. (2003) à partir de nids d’oiseaux d’une colonie mixte suggèrent la possibilité de transfert d’une espèce d’hôte à l’autre en cas de partage du lieu de nidification. Ces auteurs, en effet, notant un passage du diptère Carnus hemapterus (Carnidae) des guêpiers (Merops apiaster) aux moineaux soulcies (Petronia petronia) se contentent d'envisager une transmission similaire pour les mésostigmates hématophages aussi rencontrés (D. gallinae et O. bursa). Clayton et Tompkins (1994) considéraient la transmission de D. gallinae comme horizontale, au moins chez le pigeon, s’appuyant principalement sur le fait que l’acarien ne demeure pas sur l’hôte et sur les observations suivantes : les nids construits à proximité des nids infestés le devenaient aussi assez rapidement et des acariens avaient été vus courant sur les murs entre certains nids.
Les acariens du groupe gallinae transitent-ils par le biais des oiseaux ? Si oui, sont-ils capables de passer d'un oiseau à l'autre par simple côtoiement, en picorant sur le même terrain, ou bien un contact plus important est-il requis ? Sont-ils capables de passer d'une espèce d'hôte à l'autre et de se développer normalement ? Une fois introduit dans un nid ou dans un élevage, qu'est-ce qui préside à leur installation et à l'amplification de leurs populations ?
32 2.3 Objectifs Les objectifs de la présente étude visent à obtenir des informations quant aux voies de dissémination de l’espèce D. gallinae et aux exigences écologiques de son développement en vue, à plus long terme, de formuler des préconisations à l’intention des éleveurs. Pour cela, une approche large, tant au niveau taxinomique que méthodologique a été adoptée. En effet, préalablement à l’exploration écologique, la caractérisation précise de la ou des espèce(s) d’importance économique s’impose, et, par conséquent, une exploration élargie au-delà des espèces du groupe gallinae dans le genre Dermanyssus en vue de la délimitation précise de la ou des entités concernées. Par ailleurs, une fois l’identité spécifique de l’ (des ) acarien(s) d’importance économique clarifiée, sa spécificité d’hôte et les flux de ses populations entre volailles domestiques et oiseaux sauvages sont à fouiller si l’on veut obtenir quelques éléments d’explication aux difficultés pratiques évoquées plus haut. Pour ce faire, une appréhension comparative de la ou des espèces d’importance économique avec des espèces proches tant phylogénétiquement qu’écologiquement a été envisagée, au moyen des outils de la cladistique. Morand et al. (2002) souligne la possibilité offerte par les études phylogénétiques d’analyser des patterns et de tester des hypothèses adaptatives, en optimisant sur les topologies obtenues des informations d’ordre écologique. Dans cet esprit, une observation globale et comparative de topologies phylogénétiques impliquant diverses populations du groupe gallinae pourrait permettre d’obtenir un recul suffisant pour comprendre les mouvements des populations et certaines exigences propres aux entités posant problème en élevage. 2.4 Aperçu sommaire de l'étude : un débroussaillage en deux étapes
a - Structure du texte Le corps du texte qui suit est structuré de manière à permettre au lecteur de suivre l'évolution de l’étude représentée par six publications, avec ses rebondissements et ses réorientations successives. Le §3 présente les grandes lignes de la méthodologie adopté tout au long de l'étude. Le travail de recherche se présente ici comme une enquête dont l'étape liminaire (§4) est primordiale et sert de base pour les étapes suivantes (§5). La recherche bibliographique préalable, traditionnellement à part, est intégrée à la première partie du corps (§4, publication I), suivie de deux publications. Cet ensemble représente la longue démarche de taxinomiste pour le "débroussaillage" initial de la classification du genre Dermanyssus. La seconde partie (§5) réunit des publications contrastées aussi, mais dont la ligne conductrice est la mise à profit des résultats du "débroussaillage" pour l'exploration de certains aspects de l'écologie de cinq espèces ciblées a posteriori du genre Dermanyssus. Pour chaque publication, une présentation du contenu est sommairement développée, suivie d'éventuelles remarques visant en général à replacer certains éléments - dépassés déjà - dans le contexte actuel ou à souligner un ou des liens avec d'autres parties de la thèse.
b - Première étape (§4) : clarification de l’identité spécifique Au début de la présente étude, le genre Dermanyssus apparaissait déjà clairement défini en tant que genre par rapport aux autres genres, si l’on faisait abstraction du genre Liponyssoides (cf. § 4.3b - ). En revanche, la définition des entités spécifiques en son sein semblait fortement lacunaire (Publication I : Roy & Chauve 2007, revue). Les caractères morphologiques discriminants, traditionnellement utilisés au niveau spécifique ou à des niveaux taxinomiques supérieurs, sont dans ce genre pour la plupart fortement variables au niveau intraspécifique, voire
33 individuel (asymétries bilatérales), et les états de nombre d'entre eux sont chevauchants (Publication II : Roy & Chauve in press). Cela apparaissait particulièrement marqué chez les espèces du groupe gallinae. En effet, les espèces du groupe hirsutus, ainsi que celles du sous-genre Microdermanyssus sont porteuses de caractères morphologiques marqués (soies modifiées, aires sclérifiées plus ou moins élargies (plaque dorsale notamment), sculptées parfois de pores à la forme caractéristique) et permettant souvent la discrimination interspécifique. En revanche, les espèces du groupe gallinae – représentant 60% des espèces du genre Dermanyssus – sont faiblement sclérifiées dans l’ensemble et ne sont pas caractérisables par des éléments morphologiques nettement modifiés. Seules apparaissaient accessibles des différences subtiles dans l’agencement d’éléments dont l’observation a été standardisée pour la caractérisation de nombreux groupes d’acariens libres, à des niveaux taxinomiques variés. Mais ces différences ne semblaient pas suffisantes pour séparer les espèces dans le genre Dermanyssus. Afin d'obtenir une représentation précise de l'identité de l'espèce (ou des espèces) infestant les élevages, il est nécessaire d'appréhender l'ensemble du groupe d'espèces qui la (les) contiennent et d'y rechercher tout d'abord les limites interspécifiques. Par conséquent, il apparaissait, au début de l'étude, primordial de clarifier la délimitation entre les espèces du genre Dermanyssus, et en particulier entre les espèces du groupe gallinae, avant toute étude d'ordre écologique. Pour cela, une approche phylogénétique basée à la fois sur des caractères morphologiques et moléculaires a été choisie (Publication III Roy et al. 2009a).
c - Seconde étape (§5): exploration écologique Une fois les espèces clairement définies, au moins dans le groupe gallinae, une estimation fiable du spectre d’hôtes a été rendue possible par un échantillonnage important dans l’avifaune sauvage ainsi qu’en élevages, et l’utilisation élargie d’un des marqueurs moléculaires retenus, la mt-Co1. Une reconstruction phylogénétique d’un nombre accru de populations des espèces testées moléculairement a permis de dégager certains patterns écologiques (Publication IV Roy et al. 2009b). Ensuite, une exploration des flux de populations a pu être entamée par une approche généalogique basée sur des haplotypes (phylogéographie), complémentée par des tests de génétique des populations, grâce à l’utilisation élargie de la mt-Co1 et au développement d’un nouveau marqueur nucléaire variable intraspécifiquement. L’exploration intraspécifique basée sur ces deux loci indépendants a été réalisée de manière à comparer polymorphisme et flux de populations entre plusieurs espèces du groupe gallinae, dont D. gallinae (Publication V Roy et al, in prep). Enfin, l’influence de certaines composantes du microenvironnement (nid, litière) ont été partiellement étudiées en mettant à profit le terrain d’étude de JC Bouvier (INRA-Avignon). Des nids de mésanges (Parus sp.) ont été collectés et analysés de manière à obtenir une représentation de leur arthropodofaune en fonction de quatre modalités (nature, vergers contrôlés biologiquement, vergers contrôlés chimiquement, vergers contrôlés de façon raisonnée) et à situer l’implication du genre Dermanyssus parmi les communautés recensées (Publication VI, Roy et al, in prep). Le fil conducteur méthodologique est l'utilisation des outils de la phylogénie pour l'investigation à tous les niveaux des espèces du groupe gallinae.
34 3 Grandes lignes de la méthodologie adoptée
3.1 Matériel biologique : stratégie d’échantillonnage pour une représentation d’habitats variés Afin d'obtenir une représentation la plus complète possible des entités du genre Dermanyssus, un effort d'échantillonnage a été orienté vers différents groupes d’oiseaux, différentes origines géographiques (principalement France, mais aussi quelques échantillons transeuropéens, et nord américains) et différents types d’environnement (élevages de volailles de consommation et d'ornement, faune sauvage in natura, en ville, en agroécosystème). Pour ce faire, un réseau de vétérinaires en lien avec des élevages de volailles déjà développé à l'ENVL (Claude M. Chauve) a été mis à profit pour les prélèvements en élevage de volaille de consommation (prélèvements directs d'acariens, échantillons de litière). Pour les élevages d'ornement, des prospections auprès d'animaleries et d'associations d'éleveurs de petits oiseaux (Association Ornithologique Rhodanienne, Amicale Ornithologique Becs Crochus Centre Est (AOBC)) ont permis d'obtenir d'autres échantillons de litière. Quant à la faune sauvage, un réseau d'ornithologistes, principalement composé de bagueurs diplômés du CRBPO (Centre de Recherche sur la Biologie et les Populations d'oiseaux, Muséum National d'Histoire naturelle, 75 Paris) et d'acteurs de réserves naturelles (Ecopôle du Forez, 42, Réserve naturelle nationale des étangs du Romelaëre, Pas-de-Calais) a été développé au cours de la thèse et a permis l'obtention d'un grand nombre de nids (ou fragments de nids) d'espèces très diverses, réparties dans la France, in natura et en ville. L'implication du CSOL (Centre de Soins aux Oiseaux Sauvages du Lyonnais, Francheville), la participation à des séances de baguage d'oiseaux sauvages sous la responsabilité d'O. Caparros (CRBPO) et la participation à une journée de chasse à l'alouette avec J. Berruyer (Fédération Départementale des Chasseurs du Rhône) ont permis l'exploration de la présence sur des hôtes hors nid d'acariens du genre Dermanyssus. Des collaborateurs du CRBPO ont aussi fortement contribué à cet aspect de l'étude au cours de leurs séances respectives de baguage. Enfin, une collaboration fructueuse avec JC Bouvier (INRA- Avignon) a fourni un important échantillon de nids d'oiseaux sauvages nichant dans des agroécosystèmes de vergers de fruitiers aux caractéristiques variées. 3.2 Marqueurs développés : utilisation concomitante de données morphologiques et moléculaires Les marqueurs développés au cours de la thèse sont de deux ordres : morphologiques et moléculaires. Les données morphologiques sont toutes des caractères discrets, codés en 2 à 3 états, conformément aux aptitudes antérieurement développées dans le cadre d'une étude sur des insectes hémiptères (Roy et al. 2007). Aucun caractère continu n'a été testé ici. Le codage des caractères discrets retenus dans une matrice phylogénétique n'est utilisé que dans la première partie (délimitation interspécifique). Leur utilisation dans la seconde partie est limitée au diagnostic. Les marqueurs moléculaires reposent sur le séquençage et l'alignement de portions d'ADN génomique. En vue d'accéder à des informations complémentaires et indépendantes, des gènes du génome cytoplasmique et du génome nucléaire ont été ciblés dans le même temps. Pour la première partie de l'étude, des marqueurs couramment utilisés dans les études phylogénétiques portant sur des arthropodes, dont certains acariens (Navajas et Fenton 2000, Cruickshank 2002), ont été ciblés. Les marqueurs retenus sont situés sur deux gènes mitochondriaux (ARNr 16S, mt-Co1) et deux gènes nuclaires (fin de l'ARNr 18S- début de l'ARNr 28S - incluant les ITS1 et 2 -, elongation factor 1-D) ont été développés. Une seule 35 séquence du génome d'une espèce de Dermanyssus était disponible dans la banque de gène internationale (GenBank, EMBL) au début de l'étude. Il s'agissait d'une portion de l'ARNr 16S d' D. gallinae, utilisée par Black et Piesman (1994) comme outgroup dans une analyse des interrelations entre tiques (Acari : Ixodida). Pour ce gène, le séquençage a été obtenu par design direct d'amorces sur la séquence disponible, puis, au fil de l'étude par conception de nouvelles amorces sur les aires apparaissant conservées par alignement des premières séquences obtenues. Pour les trois autres, les amorces ont été conçues sur la base d'alignements des séquences orthologues d'autres arthropodes, disponibles sur la banque de gènes. Les trois premiers ont finalement participé à la délimitation interspécifique (publication III, p. 69 sqq.), mais le quatrième (EF1-D) a dû être abandonné, du fait d'anomalies (cf. 4.3b - ). Pour la seconde partie de l'étude, la portion de mt-Co1, non seulement porteuse de signatures spécifiques nettes, mais aussi d'une diversité intraspécifique très informative au niveau population, a été largement mise à profit. Aucun des marqueurs nucléaires développés pour la première partie n'ayant, en revanche, fourni d'information intraspécifique suffisante, un cinquième marqueur a été développé. L’utilisation d’au moins deux loci indépendants apparaissait en effet primordial pour l'exploration des flux de populations. Le marqueur nucléaire sélectionné alors consiste en deux petites portions d'exons du gène codant pour la Tropomyosine, et de l'intron qu'ils encadrent. L'ADNc complet de la Tropomyosine de D. gallinae a été publié par Nisbet et al. (2006) dans le cadre d'une étude immunologique. La fragmentation en de multiples portions de cette séquence et le criblage sur notre souche de laboratoire SK de ce gène au moyen d'amorces conçues à chacune des extrémités des fragments définis a donné lieu à l'isolement de l'amplicon retenu, à forte composante intronique. A notre connaissance, aucune analyse phylogénétique, ni de génétique des populations, n'a été menée sur la base de cette région chez des arthropodes. A la différence des reconstructions phylogénétiques de la première partie de l'étude, utilisant des régions classiquement analysées dans ce type d'étude, il s'agit ici d'une expérience originale. 3.3 Outils de la phylogénie et de la génétique des populations Les séquences obtenues ont été alignées au moyen de logiciels divers (ClustalW, MAFFT, Muscle) et dans certains cas, les alignements ont été affinés à la main (Seaview). Les matrices ainsi obtenues ont été analysées au moyen de logiciels de phylogénie, suivant différents critères. Le maximum de parcimonie avec PAUP 4.0 et TNT, ainsi que le maximum de vraisemblance avec PhyML et Phylo_win, ont été exploités directement au cours de la thèse. Grâce à la collaboration d'APG Dowling (Université de Fayetteville, Arkansas, USA), des analyses bayésiennes ont été réalisées avec le logiciel MrBayes, qui réalise une simulation technique nommée Markov chain Monte Carlo (MCMC) pour obtenir une approximation des probabilités postérieures des topologies. Les réseaux d'haplotypes, ainsi que certains outils de la génétique des populations ont été abordés en fin de thèse, dans le cadre de la seconde partie. Les outils statistiques de la génétique des populations, basés sur l'estimation de la neutralité (Nei 1987) de l'évolution trahie par les différences rencontrées dans les séquences analysées, analysent le polymorphisme de manière "statique" (à un instant t), par comparaison entre isolats, indépendamment de leur histoire. De fait, ils ne permettent pas (1) d'intégrer à l'analyse les nombreux isolats dont un petit nombre seulement d'individus avaient été séquencés pour la première partie de l'étude (statistiquement insuffisants), (2) de prendre en compte l'histoire évolutive des isolats, ce qui s'accorde difficilement avec l'esprit de l'étude dans son ensemble. Afin de réduire au moins partiellement cette frustration, des méthodes complémentaires ont été recherchées pour la publication V (p. 131 sqq.). Le logiciel Network a été mis à profit dans l'analyse des réseaux d'haplotypes. L'approche de type NCPA (Nested Clade Phylogeographic
36 Analysis), initiée par Templeton (Templeton et al. 1995, Templeton 1998), a été tout d'abord envisagée. Mais de solides arguments récemment développés contre cette approche ont découragé cette initiative (Panchal et Beaumont 2007, Petit 2007). Très attractive pour une personne habituée aux informations historiques de la cladistique et rebutée par l'aspect "monodimension" des approches statistiques, cette méthode a longtemps séduit, se présentant comme la possibilité inespérée de mettre en évidence des structures de populations en prenant en compte, précisément, leur histoire. Les simples réseaux d'haplotypes permettaient bien sûr une telle approche, mais, multidimentionnels à l'extrême par excellence au sein de l'espèce (>>3D), les relations entre entités étant typiquement réticulées du fait des flux de gène, la définition de la topologie à retenir demeurait très obscure et la robustesse des relations restait inestimée. Templeton et coll ont travaillé à l'objectivation d'une démarche d'analyse de réseaux d'haplotypes (« nichage » graduel des clades en partant des extrêmités de la topologie), suivie d'un examen statistique de la structuration des isolats testés en fonction de paramètres géographiques (ou autres) connus. Mais, si l'estimation des structures est soutenue par un volet statistique, il n'en demeure pas moins que le support de l'analyse repose sur un réseau multidimensionnel, dont la (les) topologie(s) retenue(s) l’est (le sont) sur une base obscure, plutôt subjective. Un échange fructueux avec M. Panchal (Université de Reading, Royaume Uni) a conduit au choix de méthodes plus modernes, et surtout plus objectives. M. Panchal a développé et mis au point durant sa thèse une automatisation informatique de la NCPA (demandant jusqu'alors l'utilisation successive de plusieurs logiciels indépendants) (Panchal 2007). Cette automatisation, permettant de tester un plus grand nombre de modèles, a confirmé des doutes déjà plusieurs fois formulés quant à l’efficacité de la méthode : de nombreux faux-positifs sont générés par la NCPA, induisant des conclusions fortes sur des structurations de population en fait artéfactuelles. En remplacement des analyses basées sur des réseaux d’haplotypes, M. Panchal recommande des analyses basées sur des modèles, dont l'efficacité est bien sûr encore à tester, mais dont l'objectivité apparaît nettement plus importante. Ainsi, une collaboration amorcée avec JS Lopes (Université de Reading, Royaume Uni) vise à tester les données mt-Co1 et Tropomyosine avec le logiciel PopABC. Basé sur une investigation bayésienne de la génétique des populations, ce logiciel développé par JS Lopes utilise la simulation technique Markov chain Monte Carlo (MCMC) pour l’approximation des probabilités postérieures dans le cadre de l’analyse de la démographie des populations (Lopes et Beaumont 2008). Des fourchettes quant à des paramètres démographiques concernant les populations à tester sont proposées a priori (priors) et définis de manière large. Les différents branchements possibles entre les populations prédéfinies constituent les modèles dont on teste la probabilité. Les structures sont retenues a posteriori, leur robustesse étant évaluée par les probabilités postérieures bayésiennes. Par ailleurs, une évaluation des groupements de populations a été réalisée grâce au logiciel Structure 2 et des tests statistiques classiques de génétique des populations ont été réalisés à l'aide de DnaSP v5 et Arlequin 3.11.
37
4 Taxinomie dans le genre Dermanyssus
4.1 Synthèse historique : publication I
a - Présentation Dans la nomenclature zoologique, il n’est pas rare de rencontrer des étymologies trompeuses a priori, notamment chez les acariens. Le principe du taxon type, du nom duquel découlent les noms des taxa de niveaux supérieurs, entraîne dans certains cas d’apparentes incongruités entre dénomination et habitudes écologiques de l’organisme considéré. Il suffit que le représentant du groupe concerné découvert – et donc décrit – le premier soit nommé sur la base de caractéristiques qui lui sont propres pour que la nomenclature d’ordre ou de super-familles soit jalonnées des dénominations déconcertantes. En effet, il s’avère parfois a posteriori, au moment de la description d’autres taxa, découverts ultérieurement, que lesdites caractéristiques ne soient pas communes à l’ensemble des taxa du groupe considéré. Le genre Dermanyssus, du grec « derma », peau, et « nussein », piquer, désigne à l’évidence un acarien hématophage bien davantage qu’un prédateur ou un détritivore. Or la cohorte des Dermanyssina (cf. Annexe 1), dont il représente le genre type, si elle regroupe des parasites hématophages, est originellement et majoritairement constituée de prédateurs, d’après des analyses phylogénétiques (Dowling 2006a, b) ou sur la base d’observations évolutives (Radovski 1969). Les premiers fossiles de Dermanyssina, datant de l’Eocène (Witalinski 2000), sont aussi des prédateurs, ce qui va aussi dans ce sens. La superfamille des Parasitoidea, dont le genre Parasitus est le genre type, ne regroupe pas des parasites mais des prédateurs. De même, les Dermanyssoidea ne regroupent pas seulement des ectoparasites piqueurs. Toutefois, la famille des Dermanyssidae, à l’inverse des Parasitidae qui ne comptent aucun parasite (Lesna et Sabelis, communication personnelle), ne regroupe aujourd’hui que des ectoparasites hématophages. Cette famille a en revanche longtemps regroupé des acariens aux habitudes diverses et est caractérisée par une histoire chaotique. Elle a tout d’abord représenté la famille la plus importante (en nombre d’espèces incluses) des Dermanyssoidea. Evans & Till (1966) comptaient dans les Dermanyssidae 15 sous-familles contenant de nombreux hématophages, mais pas seulement. Les Laelapinae, qui représentent la plus grande sous-famille, demeuraient principalement composée de prédateurs. La famille des Dermanyssidae s’est vu réduire drastiquement à deux genres en 1966 par Radovski : Dermanyssus et Liponyssoides. Vidée de cette pléthore de genres, elle apparaît minuscule maintenant, comparée aux familles apparentées, qui ne sont autres que les sous-familles qui la constituaient, sans les Dermanyssinae. Mais les arguments de Radovski, basés sur une analyse évolutive, si ce n’est phylogénétique, sont solides, puisqu’ils ont permis, enfin, une délimitation claire des Dermanyssidae par rapport aux autres familles. Aucun autre genre que Dermanyssus et Liponyssoides n’a été intégré depuis lors dans cette famille. Les Macronyssidae, en revanche, contiennent aujourd’hui 26 genres, les Laelapidae 144, les Haemogamasidae 7, les Hirstionyssidae 5, … (selon Hallan 2005). Parallèlement, le genre Dermanyssus a contenu au moins 60 espèces différentes. Dugès l’a décrit en 1834 sur la base de D. gallinae, l’espèce-type et le problématique pou rouge des volailles. Il lui a alors associé 4 autres espèces, dont une, D. convolvuli (littéralement « du liseron ») représente probablement une espèce non hématophage et prédatrice d’acariformes phytophages. Il est probable que D. gallinae représentait alors déjà depuis longtemps un problème important dans les élevages, puisqu’il a été décrit précocement (en 1778, dans le genre Acarus, genre type des acariens) par rapport à des Dermanyssoidea prédateurs, comme certains Laelapidae, Ascidae ou
38 Macrochelidae. Lesdits prédateurs, de taille comparable au pou rouge pourtant, mais nettement moins gênants, attirent de fait beaucoup moins l’attention. Ainsi les déprédations produites par le pou rouge sont-elles probablement la cause de l’omniprésence du radical dermanyss- dans la nomenclature d’un groupe au cours de l’évolution duquel les parasites piqueurs ne sont apparus que sporadiquement (Dowling 2006a).
4.1.a.1 Objectifs L’objectif principal de la revue historique était de faire le point sur la composition taxinomique du genre Dermanyssus au niveau spécifique, l’état des délimitations interspécifiques et sa caractérisation au niveau générique en 2006. Pour cela, un examen précis de l’histoire de sa taxinomie depuis sa description et des caractérisations morphologiques utilisées dans la littérature a été envisagé. Dans un but pratique, un aperçu de la répartition géographique et des spectres d’hôte a été intégré à l’article.
4.1.a.2 Principaux résultats Au moment de la revue, Dermanyssus ne regroupait plus que 23 espèces. Une discussion entre Krantz et Sheals, ainsi que la révision du genre par Evans et Till en 1962 ont contribué à clarifier la définition du genre par rapport aux autres genres de Dermanyssoidea. Cette dernière constitue la première révision complète du genre Dermanyssus, plus d’un siècle après sa description, et a drastiquement réduit le nombre d’espèces en son sein (14 espèces). Elle fut complétée ensuite par les travaux de Moss (1967, 1968, 1978 : 18 espèces), qui introduisit des subdivisions et participa à la définition des espèces au sein du genre. Mais Moss mit en évidence en 1978, avec l’introduction dans son analyse de nouvelles espèces, le caractère inapproprié d’une partie de ses subdivisions. Malheureusement, il ne publia jamais la nouvelle étude annoncée. Plusieurs auteurs ont dénoncé, au fil du temps, l’extrême variabilitié intraspécifique de certains des caractères morphologiques proposés pour la définition des espèces et le diagnostic.
En outre, la claire délimitation entre Dermanyssus et Liponyssoides n’est pas apparue si tranchée que Moss ne le suggèrait en 1967. Là aussi, Moss annonçait une révision du genre Liponyssoides, qu’il ne publia jamais. Les deux genres de la famille des Dermanyssidae se différencient nettement des autres genres de Dermanyssoidea par la structure particulière de leurs chélicères, fortement modifiée par l’adaptation à l’hématophagie (et de manière différente d’autres Dermanyssoidea hématophages tels les Macronyssidae), la texture souple des cornicules et la pointe sur les tarses III et IV chez le mâle. En revanche, la distinction Dermanyssus / Liponyssoides, principalement basée sur la forme de la plaque sternale ainsi que certains éléments de chaetotaxie, tous fortement polymorphes au sein de populations, et même des individus, demeurait (et demeure) peu nette.
Une répartition mondiale et une distribution par espèce d’hôte recensée, sur la base des données de la littérature, ont aussi été fournies.
b - Remarques sur la publication I
4.1.b.1 Nombre d’espèce augmenté (2008 et 2009) Au moment de la publication I, 23 espèces étaient incluses dans le genre Dermanyssus. Aujourd'hui, 25 espèces sont à prendre en compte : l'Américain W. Knee a décrit D. diphyes en 2008 et nous avons nous-mêmes décrit D. apodis en 2009 (Publication III, p. 69 sqq.).
39 PUBLICATION I
HISTORICAL REVIEW OF THE GENUS DERMANYSSUS DUGÈS, 1834 (ACARI: MESOSTIGMATA: DERMANYSSIDAE) ROY L.* & CHAUVE C.M.*
Summary: Résumé : REVUE HISTORIQUE DU GENRE DERMANYSSUS (ACARI : MESOSTIGMATA : DERMANYSSIDAE) A synthetic review of the historical systematics of Dermanyssus Dugès, 1834 (Acari: Mesostigmata: Dermanyssidae) is provided. La systématique historique de Dermanyssus Dugès, 1834 (Acari : The classification at the specific level in this early genus has not Mesostigmata : Dermanyssidae) est révisée de manière really been clarified during more than a century despite its synthétique. La classification au niveau spécifique de ce vieux economic impact, and the history of the genus is complex and genre n’a pas été véritablement clarifiée durant plus d’un siècle includes various stages. Moreover, Dermanyssus currently includes malgré son impact économique, si bien que son histoire est 23 species, whereas the last review took only 18 species into quelque peu complexe et présente diverses étapes. En outre, le account. Changes in the species status and position in the genus genre Dermanyssus englobe à l’heure actuelle 23 espèces, tandis Dermanyssus from 1834 until today are presented. The evolution que la dernière révision prenait seulement 18 espèces en compte. of the generic definition is explored and compared with other Les changements de statut et de position des espèces dans le genera of the group. How the discrimination between the different genre Dermanyssus depuis 1834 jusqu’à présent sont présentés. species evolved in the genus is also examined. Some difficulties in Le processus d’évolution de la définition du genre est exploré par the specific definitions are discussed. A current diagnosis of the comparaison avec d’autres genres du groupe. La manière dont la genus Dermanyssus is given. A table of the species included in discrimination entre les différentes espèces a évolué est aussi this genus since its first description along with their respective considérée. Certaines difficultés dans la définition spécifique sont current positions, a list of the currently included species in discutées. Une description diagnostique actualisée du genre Dermanyssus with their hosts, and a world map presenting their Dermanyssus est fournie. Un tableau des espèces qui ont été geographic distribution are provided. incluses dans le genre depuis sa création avec leur position respective, une liste des espèces actuelles du genre Dermanyssus KEY WORDS : Acari, Mesostigmata, Gamasida, Dermanyssus, historical avec leurs hôtes, ainsi qu’une carte de leur distribution mondiale review, systematics. sont fournis.
MOTS CLÉS : Acari, Mesostigmata, Gamasida, Dermanyssus, revue historique, systématique.
INTRODUCTION group name, including only obligatory ectoparasites. Radovsky (1966, 1967) separated this group into the two he genus Dermanyssus Dugès, 1834 (Acari: Meso- following families, depending on some morphological stigmata: Dermanyssidae) includes hematopha- and biological characters: Macronyssidae Oudemans, Tgous mite species which are ectoparasites of 1936 and Dermanyssidae. Really Dermanyssidae appear birds. Dermanyssus is the type genus of a family whose phylogenetically closer to free-living Laelapids than to name has represented various groups all along 19th cen- Macronyssidae. Consequently, Moss (1968, 1978) consi- tury, with more or less internal splitting. Dermanyssidae dered only the two following genera to be included in Kolenati, 1859 included first mites with diverse habits, Dermanyssidae: Dermanyssus and Liponyssoides. In this some of them being obligatory ectoparasites and others paper, we follow this last classification. free-living or facultative ectoparasites. Berlese (1892) Along with Liponyssoides, Dermanyssus possesses some separated the former and the last group in two diffe- morphological mainly located in the mouthparts adap- rent families: Dermanyssidae and Laelapidae respecti- tations to hematophagic habits. Adult females, proto- vely. Then numerous steps occurred, Dermanyssidae nymphs and deutonymphs possess conspicuously thin status alternating from subfamily-group name to family- and elongated chelicerae, with a second segment adap- ted to hematophagy. Faces of opposed second seg- * Laboratoire de Parasitologie et Maladies Parasitaires, École Natio- ments are medially concave, so that they may form a nale Vétérinaire de Lyon, 1, avenue Bourgelat, 69280 Marcy-L’Étoile, tube by joining together through which blood is with- France. drawn. Chelae are conspicuously reduced, even if Correspondence: Lise Roy. Tel.: 00 33 4 78 87 25 74 – Fax: 00 33 4 78 87 25 77. digits can be seen with a scanning electron microscope E-mail : [email protected] (cf. Fig. 2; Radovsky, 1969; Phillis, 2006). Male cheli-
Parasite, 2007, 14, 87-100 Mise au point 87 PUBLICATION I ROY L. & CHAUVE C.M.
D. alaudae 4 D. trochilinis 5 D. Triscutatus D. transvaalensis D. rwandae D. quintus D. prognephilus D. passerinus D. nipponensis D. longipes D. hirundinis D. hirsutus D. grochovskae D. gallinoides D. faralloni D. chelidonis D. carpathicus D. brevis D. brevirivulus ) D. antillarum D. americanus
D. wutaiensis A "
B C
Fig. 1. – Distribution map of non-gallinae species of Derma- nyssus. A: Europe. B: Asia. C: Africa. D: America. Note: a symbol corresponds to the presence of one species in a country (admi- nistrative).
D
Parasite, 2007, 14, 87-100 88 Mise au point PUBLICATION I
HISTORICAL REVIEW OF THE GENUS DERMANYSSUS
cerae are broader and male chelae are enlarged, with many species of Dermanyssus are distributed on more a long spermadactyl. than one continent (Fig. 1). The history of Dermanyssus The poultry red mite D. gallinae (De Geer, 1778) is is very complicated and has never been extensively exa- very common in layer houses in Europe. The economic mined. Moreover, this early genus currently includes impact of this parasite is quite important and may take 23 species, whereas the last review of the genus took many forms, including the following: downgraded only 18 species into account. For both these reasons eggs, decreased egg production, anaemia, possible and in order to get a clear view of the genus before death from exsanguination. The poultry red mite can reviewing it, it seemed necessary to examine it curso- also transmit diseases such as avian spirocheatosis, fowl rily from its description until the present and to check cholera, salmonellosis, etc. Despite such economic the current species included in it. In order to get a view importance, the classification of this genus at species of the generic history, the text will be broken down as level has been in a state of confusion for many years. follows: changes in the species status and position in About 40-50 years ago, some authors began working the genus Dermanyssus from 1834 until today are pre- precisely on the genus Dermanyssus. According to sented first, then the generic definition and its evolu- some of them, D. gallinae may not be the only Der- tion are explored. Afterwards, the species definitions manyssus parasitising laying hens. Consequently, some and their difficulties will be examined. Finally, concer- other closely related species might often have been ning the genus as is currently defined, a list of the cur- confused with this species. A rather low host-specifi- rently included species in Dermanyssus is provided. city and a rather wide geographic distribution of Der- Abbreviations: Setal terminology follows Lindquist & manyssus species contribute to obscure the issue. Evans, 1965 for the dorsum and Evans, 1963 for the legs. Most species of this genus are not very host specific: for instance, more than 30 bird species are potential hosts for D. gallinae (Zemskaya, 1971) and 40 bird spe- CHANGES IN THE STATUS AND cies (belonging to eight different orders) for D. hirun- dinis (Hermann, 1804) (Moss et al., 1970; Moss, 1978; THE POSITION OF SPECIES IN THE GENUS Fend’a & Schniererová, 2004; Fend’a, unpublished data). DERMANYSSUS FROM 1834 UNTIL TODAY Most of the Dermanyssus species are nidicolous. Although some of them can be found frequently on the ugès described the genus Dermanyssus in 1834, host and can deposit their eggs on its feathers (D. gro- in which he included five new species: D. avium chovskae Zemskaya, 1961, D. quintus Vitzthum, 1921 DDugès, 1834, D. vespertilionis Dugès, 1834, and D. americanus Ewing, 1922), most of them climb D. convolvuli Dugès, 1834, D. oribatis Dugès, 1834, and onto their host only to get a meal and then go back D. hominis Dugès, 1834. The type-species D. gallinae to their hiding-place in the host nest or roost. Moreover, (De Geer, 1778) was described by De Geer in the
Fig. 2. – Left: characte- ristically reduced chela in D. gallinae. Right: se- cond segments of cheli- cerae medially concave in D. gallinae (electron scanning microscope).
Parasite, 2007, 14, 87-100 Mise au point 89 PUBLICATION I ROY L. & CHAUVE C.M.
genus Acarus. In 1834, Dugès named it D. avium, and lvuli, Dugès only noted for each: species name, fol- considered A. gallinae, although senior, synonymous lowed by a comma and the personal latin pronoun with D. avium. D. gallinae was later reinstated as the nobis. Without any description and as species names senior synonym (Koch, 1836). D. vespertilionis has been suggest host associations far from the common ones, suppressed by the International Commission of Zoo- birds, in Dermanyssus, D. oribatis and D. convolvuli logical Nomenclature (ICZN) under the plenary powers might be deemed nomina dubia. In any case, they for the principle of priority, but not for homonymy cannot be included in Dermanyssidae sensu Radovsky (Melville & Smith, 1987). About D. oribatis and D. convo- (1966). D. hominis seems to have been omitted by all
Species included or previously included in Dermanyssus Current position Comments
1778 * D. gallinae (De Geer, 1778) Dermanyssidae Dermanyssus * D. alaudae (Schrank, 1781) Dermanyssidae Dermanyssus In 1781, Schrank named alaudae the seventh species he described in 1776. * D. hirundinis (Hermann, 1804) Dermanyssidae Dermanyssus D. truncatus (Olfers, 1816) synonymy ➣ D. alaudae D. hominis (Dugès, 1834) synonymy ➣ D. gallinae Bory de Saint Vincent described this species in 1823 and more completely in 1828 without any name nor systematic position within the Acari group. Dugès placed it in genus Dermanyssus and named it in 1834. D. avium Dugès, 1834 synonymy ➣ D. gallinae D. vespertilionis Dugès, 1834 suppressed ICZN direction 66: suppressed under the plenary powers for the principle of priority, but not for homonymy. D. convolvuli Dugès, 1834 ? D. oribatis Dugès, 1834 ? D. musculi Koch, 1836 Macronyssidae Steatonyssus This species has complicated history. Oudemans (1936) considered it junior synonym of A. musculi Shrank, which he placed in genus Steatonyssus (homo- nymy and synonymy in the same time). Evans and Till (1966:278-279) suggested that S. musculi Schrank could be a junior synonym of Ornithonyssus bacoti Hirst, 1913. D. arcuatus Koch, 1839 Hirstionyssidae Echinonyssus D. carnifex Koch, 1839 Hirstionyssidae Echinonyssus Tenorio (1984) treats this species as nomen dubium. D. coriaceus Koch, 1839 synonymy ➣ D. arcuatus D. lanius Koch, 1839 synonymy ➣ D. carnifex D. noctulae Koch, 1839 synonymy ➣ D. arcuatus D. murinus Lucas, 1840 Macronyssidae Steatonyssus D. avium Wagner, 1841 synonymy ➣ D. murinus D. pipistrellae Koch, 1841 synonymy ➣ D. arcuatus D. lacertarum (Contarini, 1843) ? D. natricis Gervais, 1844 Macronyssidae Ophionyssus D. musculi Johnston, 1849 Hirstionyssidae Echinonyssus A complicated history. Seems to be conspecific to D. musculi Koch which is conspecific to A. musculi Shrank. D. flavus Kolenati, 1857 Macronyssidae Macronyssus D. glutinosus Kolenati, 1857 synonymy ➣ M. granulosus D. granulosus Kolenati, 1857 Macronyssidae Macronyssus D. ambulans Thorell, 1872 Haemogamasidae Haemogamasus D. richiardii Canestrini & Fanzago, 1877 ? D. sylviarum Canestrini & Fanzago, 1877 Macronyssidae Ornithonyssus D. hirundinis Berlese, 1889 homonymy ➣ nomen novum: D. chelidonis * D. longipes Berlese &Trouessart, 1889 Dermanyssidae Dermanyssus nomen dubium. 1889 * D. passerinus Berlese & Trouessart, 1889 Dermanyssidae Dermanyssus species inquirenda.
Table I (to be continued).
Parasite, 2007, 14, 87-100 90 Mise au point PUBLICATION I
HISTORICAL REVIEW OF THE GENUS DERMANYSSUS
Species included or previously included in Dermanyssus Current position Comments
1902 D. albatus Oudemans, 1902 synonymy ➣ D. arcuatus D. aegyptius Hirst, 1913 Dermanyssidae Liponyssoides D. muris Hirst, 1913 Dermanyssidae Liponyssoides D. sanguineus Hirst, 1914 Dermanyssidae Liponyssoides * D. quintus Vitzthum, 1921 Dermanyssidae Dermanyssus * D. americanus Ewing, 1922 Dermanyssidae Dermanyssus D. oti Ewing, 1925 synonymy ➣ D. americanus D. evotomydis Ewing, 1933 synonymy ➣ D. gallinae * D. prognephilus Ewing, 1933 Dermanyssidae Dermanyssus D. brasiliensis Fonseca, 1935 Dermanyssidae Liponyssoides * D. brevis Ewing, 1936 Dermanyssidae Dermanyssus D. scutatus Ewing, 1936 homonymy ➣ nomen novum: D. hirsutus * D. chelidonis Oudemans, 1939 Dermanyssidae Dermanyssus This species has been described in 1889 by Berlese as D. hirundinis. Because of homonymy, Oudemans renamed it in 1939. * D. triscutatus Krantz, 1959 Dermanyssidae Dermanyssus * D. grochovskae Zemskaya, 1961 Dermanyssidae Dermanyssus * D. transvaalensis Evans & Till, 1962 Dermanyssidae Dermanyssus D. intermedius Evans & Till, 1964 Dermanyssidae Liponyssoides * D. gallinoides Moss, 1966 Dermanyssidae Dermanyssus * D. faralloni Nelson & Furman, 1967 Dermanyssidae Dermanyssus * D. hirsutus Moss & Radovsky, 1967 Dermanyssidae Dermanyssus * D. antillarum Dusbabek & Cerny, 1971 Dermanyssidae Dermanyssus * D. trochilinis Moss, 1978 Dermanyssidae Dermanyssus * D. carpathicus Zeman, 1979 Dermanyssidae Dermanyssus * D. nipponensis Uchikawa & Kitaoka, 1981 Dermanyssidae Dermanyssus * D. brevirivulus Gu & Ting, 1992 Dermanyssidae Dermanyssus * D. wutaiensis Gu & Ting, 1992 Dermanyssidae Dermanyssus 1993 * D. rwandae Fain, 1993 Dermanyssidae Dermanyssus
Table I. – Species included or previously included in Dermanyssus listed in chronological order with their present position/status. Spe- cies names preceded by * are here included in Dermanyssus. reviewers of Dermanyssus until today (cf. § Species Ewing, 1936, D. scutatus Ewing, 1936, D. chelidonis whose nomenclatural status is not clear). To sum up, Oudemans, 1939. only one of the five initially included species has been Evans & Till (1962) recognized 14 species, two of considered in the subsequent studies. which, overlooked by Strandtmann & Wharton (1958), After these descriptions, many species were created in were considered doubtful but not to be invalidated Dermanyssus by other authors. We list 57 species which (D. passerinus Berlese & Trouessart, 1889 and D. lon- are included or have been included in the genus gipes Berlese & Trouessart, 1889). Two others had been (Table I). 32 species are changed in status or position: described after 1958 (D. triscutatus Krantz, 1959 and ten species have been synonymized with other Der- D. grochovskae), one was a new species (D. trans- manyssus species. Two species receive nomina nova vaalensis Evans & Till, 1962). Another one, which had because they are deemed junior homonyms. One was been considered synonymous with D. gallinae by suppressed under the plenary powers (cf. above). 16 spe- Oudemans, was restored (D. alaudae (Schrank, 1781)). cies are now included in some other groups: five in From the ten listed species in Strandtmann and Whar- the other genus of Dermanyssidae Liponyssoides, seven ton 1958, two were synonymized (D. oti with D. ame- in several genera of the family Macronyssidae, three ricanus and D. evotomydis with D. gallinae). in the family Hirstionyssidae, one in the family Hae- Moss (1968) also included 14 species in Dermanyssus, mogamasidae. Four are incertae sedis or species inqui- but not exactly the same as Evans & Till (1962). The renda (cf. § Species whose nomenclatural status is not differences were: the newly named D. hirsutus Moss clear). Finally, one species is suggested here being & Radovsky, 1967 (= D. scutatus, praeocc.), the recently synonymized (D. hominis). As a result, 23 species are described D. gallinoides Moss, 1966 and D. faralloni included in Dermanyssus. Nelson & Furman, 1967 (in a footnote, because this Strandtmann & Wharton (1958) listed ten species in species had been described as Moss’s paper went to Dermanyssus: D. gallinae, D. hirundinis, D. quintus, press), and the omission of D. passerinus and D. lon- D. americanus, D. oti Ewing, 1925, D. evotomydis gipes. In 1978, Moss added four species: D. antillarum Ewing, 1933, D. prognephilus Ewing, 1933, D. brevis Dusbabek & Cerny, 1971, D. trochilinis Moss, 1978,
Parasite, 2007, 14, 87-100 Mise au point 91 PUBLICATION I ROY L. & CHAUVE C.M.
D. passerinus and D. longipes, the last two being consi- deeper investigation of the description of Derma- dered incertae sedis and not included in the key for nyssus. Krantz described the first Dermanyssus species identification. Thus, in the last review of genus Der- having a divided dorsal shield in the adult stage: D. tris- manyssus, only 18 species were included. cutatus (dorsal shield short, several metanotal scutella The five species which have been described after Moss’s present). He also pointed out the fact that the discri- last review are: D. brevirivulus Gu & Ting, 1992, D. car- mination between Allodermanyssus and Dermanyssus pathicus Zeman, 1979, D. nipponensis Uchikawa & based on the character incomplete/complete dorsal Kitaoka, 1981, D. rwandae Fain, 1993, D. wutaiensis Gu shield in the adult stage is not correct anymore. But & Ting, 1992. he neglected to consider the genus Liponyssoides. Sheals examined the three genera together. In order to explain the new synonymy he established between Alloder- EVOLUTION OF THE GENUS DEFINITION manyssus and Liponyssoides, he provided some argu- COMPARED WITH THE OTHER GENERA ments concerning the ontogeny (one seta less on the femur and one less on the palp genu in the adult stage OF THE GROUP in Dermanyssus). Another apparently important argument concerned the his early genus definition follows a somewhat chaetotaxy of the dorsal shield and applied, according complex evolution, and it is necessary to to Sheals, not only to Dermanyssinae but also to Macro- Texplore its history throughout the literature in nyssinae (today Macronyssidae; cf. infra): he considered order to understand it. Dugès (1834) described the the presence/absence of seta j3 a character related to genus Dermanyssus as follows: “Palporum articulus the host group in both taxa (present in all parasites of us 5 minimus; labium acutum; mandibulae maribus mammalian and absent in all parasites of birds). How- chelatae, ungue longissimo, feminis ensiformes; corpus ever, it should be noted that the more recently des- molle; pedes antici longiore; coxae contiguae. Larvae cribed species D. trochilinis is an exception to such a hexapodae, adultis vix dissimiles”. Such a morpholo- hypothesis: it is parasitic on birds and doesn’t lack j3. gical description appears today extremely general and Moreover, genus Ornithonyssus (Macronyssidae) lacks fits most current mesostigmates. About a century later, j3 and includes species which are parasitic on mam- in 1923 and in 1936, Ewing provided short surveys of mals (e.g. O. bacoti on rodents). In short, Sheals was this genus in North America. The first one, included wrong in this last hypothesis. in a review of North American dermanyssids, listed Secondly, Evans & Till (1962) stabilized Dermanyssus only two species, and the second survey, being a com- description: they wrote the first worldwide monograph pact summation, included several recently described on the genus Dermanyssus. Many generic characters species. were based on the ontogeny and the chaetotaxy of In 1958, Strandtmann and Wharton, in a large opus shields and legs. reviewing the classification of the mesostigmates para- Finally, Radovsky (1966, 1967) established Macro- sitic on vertebrates, pointed out the serious need of nyssinae (Mesostigmata: Dermanyssidae) as the revision of the genus Dermanyssus: “The genus is in macronyssid family. This status change is very impor- need of revision. It is doubtful that all the species listed tant. From then on, dermanyssids species i.e. Der- below really are specific entities” (Strandtmann & manyssus spp. and Liponyssoides spp., are to be dis- Wharton, 1958, p. 122). tinguished from macronyssids in having the 2nd From then on, three steps can be distinguished, on the segment of the chelicerae elongate and very slender whole, which lead to the current and stabilized des- (the 1st one is elongate, and differently conformed, cription. Many other genera have been created, which in Macronyssids), the chelae reduced (edentate, but are more or less closely related to Dermanyssus. each digit visible with an optic miscroscope in Macro- First, two of these genera are very closely related to nyssids) and a deutonymphal stage which needs a Dermanyssus: Allodermanyssus Ewing, 1923 and Lipo- blood meal in order to accomplish its moulting (deu- nyssoides Hirst, 1913. The exploration of both these tonymphs moult without feeding, as do larvae, in genera compared with Dermanyssus helped the defi- macronyssids). nition to become more precise. Krantz (1959) and Sheals (1962) took part in the evolution of the genus defini- CURRENT DIAGNOSIS OF DERMANYSSIDAE tion, discussing the relationships among the three clo- sely related genera. According to both the authors, Dermanyssids are characterized among Dermanyssoi- Allodermanyssus was not valid anymore. But Krantz consi- dea by the following characters: dered Allodermanyssus synonymous with Dermanyssus, . Adult females whereas Sheals considered Allodermanyssus synony- Gnathosoma-chelicerae: distal segment (= 2nd) of the mous with Liponyssoides. This discordance induced a female chelicerae conspicuously elongated and slender,
Parasite, 2007, 14, 87-100 92 Mise au point PUBLICATION I
HISTORICAL REVIEW OF THE GENUS DERMANYSSUS
Dermanyssus species Host species
D. alaudae (Schrank, 1781) Alauda arvensis Alauda: Alaudidae: Passeriformes Lululla arborea Alauda: Alaudidae: Passeriformes D. americanus Ewing, 1922 Carpodacus lexicanus Carpodacus: Fringilloidea: Passeriformes Emberiza cioides Emberiza: Fringillidae: Passeriformes Otus asio Otus: Strigidae: Strigiformes Passer domesticus Passer: Passeridae: Passeriformes P. montanus Passer: Passeridae: Passeriformes Serinus canaries Serinus: Fringillidae: Passeriformes Sitta sp. Sitta: Sittidae: Passeriformes D. antillarum Dusbabek & Cerny, 1971 Mimus polyglottos orpheus Mimus: Sturnidae: Passeriformes Passer domesticus Passer: Passeridae: Passeriformes Accipiter striatus fringilloides Accipiter: Accipitridae: Ciconiiformes Tachornis phoenicobia Tachornis: Apodidae: Apodiformes D. brevirivulus Gu & Ting, 1992 Galerida cristata leautungensis Galerida: Alaudidae: Passeriformes D. brevis Ewing, 1936 Alauda arvensis Alauda: Alaudidae: Passeriformes Eremophila alpestris Eremophila: Alaudidae: Passeriformes D. carpathicus Zeman, 1979 Phoenicurus phoenicurus Phoenicurus: Muscicapidae: Passeriformes Parus major Parus: Paridae: Passeriformes D. chelidonis Oudemans, 1939 Carduelis carduelis Carduelis: Fringillidae: Passeriformes Delichon urbica Delichon: Hirundinidae: Passeriformes Hirundo rustica Hirundo: Hirundinidae: Passeriformes Parus coeruleus Parus: Paridae: Passeriformes Riparia riparia Riparia: Hirundinidae: Passeriformes Ptyonoprogne rupestris Ptyonoprogne = Hirundo: Hirundinidae: Passeriformes D. faralloni Nelson & Furman, 1967 Oceanodroma homochroa Oceanodroma: Hydrobatidae: Ciconiiformes Cepphus columba Cepphus: Alcidae: Ciconiiformes Ptychoramphus aleutica Ptychoramphus: Alcidae: Ciconiiformes D. gallinae (De Geer, 1778) Acrocephalus arundinaceus Acrocephalus: Sylviidae: Passeriformes Aegolius funereus Aegolius: Strigidae: Strigiformes Carduelis carduelis Carduelis: Fringillidae: Passeriformes Carduelis spinus Carduelis: Fringillidae: Passeriformes Columba livia Columba: Columbidae: Columbiformes Delichon urbica Delichon: Hirundinidae: Passeriformes Emberiza citrinella Emberiza: Fringillidae: Passeriformes Erithacus rubecula Erithacus: Muscicapidae: Passeriformes Ficedula albicollis Ficedula: Muscicapidae: Passeriformes Ficedula hypoleuca Ficedula: Muscicapidae: Passeriformes Hirundo rustica Hirundo: Hirundinidae: Passeriformes Jynx torquilla Jynx: Picidae: Piciformes Merops apiaster Merops: Meropidae: Coraciiformes Parus major Parus: Paridae: Passeriformes P. ater Parus: Paridae: Passeriformes Passer domesticus Passer: Passeridae: Passeriformes P. montanus Passer: Passeridae: Passeriformes Phoenicurus phoenicurus Phoenicurus: Muscicapidae: Passeriformes Remiz pendulinus Remiz: Paridae: Passeriformes Riparia riparia Riparia: Hirundinidae: Passeriformes Serinus canarius Serinus: Fringillidae: Passeriformes Sitta europaea Sitta: Sittidae: Passeriformes Sturnus vulgaris Sturnus: Sturnidae: Passeriformes Other wild birds and numerous Galliformes, Anseriformes species of domestic fowl, etc. Sometimes on mammalian species (insectivora, rodents, man) D. gallinoides Moss, 1966 Asyndesmus lewis Asyndesmus = Melanerpes: Picidae: Piciformes Colaptes cafer (= C. auratus) Colaptes: Picidae: Piciformes Dendrocopos pubescens Dendrocopos: Picidae: Piciformes Dryocopus pileatus Dryocopus: Picidae: Piciformes Sphyrapicus varius Sphyrapicus: Picidae: Piciformes
Table II (to be continued).
Parasite, 2007, 14, 87-100 Mise au point 93 PUBLICATION I
ROY L. & CHAUVE C.M.
Dermanyssus species Host species
D. grochovskae Zemskaya, 1961 Nucifraga caryocatactes Nucifraga: Corvidae: Passeriformes Dendrocopos leucotos Dendrocopos: Picidae: Piciformes D. major Dendrocopos: Picidae: Piciformes Picus awokera awokera Picus: Picidae: Piciformes D. hirsutus Moss & Radovsky, 1967 Colaptes cafer (= C. auratus) Colaptes: Picidae: Piciformes D. hirundinis (Hermann, 1804) Acrocephalus arundinaceus Acrocephalus: Sylviidae: Passeriformes A. palustris Acrocephalus: Sylviidae: Passeriformes A. scirpaceus Acrocephalus: Sylviidae: Passeriformes Anthus arboreus Anthus: Motacillidae: Passeriformes Apus affinis Apus: Apodidae: Apodiformes Aquila pomarina Aquila: Accipitridae: Passeriformes Anser anser Anser: Anatidae: Anseriformes Aythya fuligula Aythya: Anatidae: Anseriformes A. ferina Aythya: Anatidae: Anseriformes Chaetura pelagica Chaetura: Apodidae: Apodiformes Columba livia Columba: Columbidae: Columbiformes Delichon urbica Delichon: Hirundinidae: Passeriformes Dendrocopos pubescens Dendrocopos: Picidae: Piciformes Ficedula albicollis Ficedula: Muscicapidae: Passeriformes Hirundo rustica Hirundo: Hirundinidae: Passeriformes H. urbica Hirundo: Hirundinidae: Passeriformes Iridoprocne bicolor Iridoprocne = Tachynecita: Hirundinidae: Passeriformes Lanius minor Lanius: Laniidae: Passeriformes L. collurio Lanius: Laniidae: Passeriformes Luscinia megarhynchos Luscinia : Muscicapidae: Passeriformes Merops apiaster Merops: Meropidae: Coraciiformes Micropus affinis Micropus = Apus: Apodidae: Apodiformes Parus caeruleus Parus: Paridae: Passeriformes P. major Parus: Paridae: Passeriformes P. palustris Parus: Paridae: Passeriformes Passer montanus Passer: Passeridae: Passeriformes P. domesticus Passer: Passeridae: Passeriformes Petrochelidon pyrrhonota Petrochelidon: Hirundinidae: Passeriformes Phoenicurus ochruros Phoenicurus: Muscicapidae: Passeriformes Remiz pendulinus Remiz: Paridae: Passeriformes Riparia riparia Riparia: Hirundinidae: Passeriformes Sitta europaea Sitta: Sittidae: Passeriformes Strix aluco Strix: Strigidae: Strigiformes Sturnus vulgaris Sturnus: Sturnidae: Sturniformes Taeniopygia guttata castanotis Taeniopygia: Passeridae: Passeriformes Troglodytes aedon Troglodytes: Certhiidae: Passeriformes T. troglodytes Troglodytes: Certhiidae: Passeriformes Turdus torquatus Turdus: Muscicapidae: Passeriformes Vireo olivaceus Vireo: Vireonidae: Passeriformes Sometimes on mammalian species (insectivora, rodents) D. longipes Berlese & Trouessart, 1889 Passer domesticus Passer: Passeridae: Passeriformes (incertae sedis) D. nipponensis Uchikawa & Kitaoka, 1981 Picus awokera awokera Picus: Picidae: Piciformes D. passerinus Berlese & Trouessart, 1889 Emberiza cirtus Emberiza: Fringillidae: Passeriformes (incertae sedis) Ficedula albicollis Ficedula: Muscicapidae: Passeriformes Jynx torquilla Jynx: Picidae: Piciformes Parus major Parus: Paridae: Passeriformes Passer italiae Passer: Passeridae: Passeriformes Sturnus vulgaris Sturnus: Sturnidae: Sturniformes D. prognephilus Ewing, 1933 Progne subis subis Progne: Hirundinidae: Passeriformes Colaptes cafer (= C. auratus) Colaptes: Picidae: Piciformes Dendrocopos pubescens Dendrocopos: Picidae: Piciformes Melanerpes erythrocephalus Melanerpes: Picidae: Piciformes Molothrus ater Molothrus: Fringillidae: Passeriformes Sialia sialis Sialia: Muscicapidae: Passeriformes
Table II (to be continued).
Parasite, 2007, 14, 87-100 94 Mise au point PUBLICATION I
HISTORICAL REVIEW OF THE GENUS DERMANYSSUS
Dermanyssus species Host species
D. quintus Vitzthum, 1921 Dendrocopos major Dendrocopos: Picidae: Piciformes D. pubescens Dendrocopos: Picidae: Piciformes Dryobates leucotes Dendrocopos: Picidae: Piciformes D. major Dendrocopos: Picidae: Piciformes Picoides pubescens Picoides: Picidae: Piciformes P. tridactylus Picoides: Picidae: Piciformes Picus viridis Picus: Picidae: Piciformes D. rwandae Fain, 1993 Apus affinis Apus: Apodidae: Apodiformes D. transvaalensis Evans & Till, 1962 Hirundo spilodera Hirundo: Hirundinidae: Passeriformes Petrochelidon spilodera Petrochelidon: Hirundinidae: Passeriformes D. triscutatus Krantz, 1959 Hirundo sp. Hirundo: Hirundinidae: Passeriformes Petrochelidon pyrrhonota Petrochelidon: Hirundinidae: Passeriformes D. trochilinis Moss 1978 Trochilidae Trochilidae: Apodiformes D. wutaiensis Gu & Ting, 1992 Passer montanus Passer: Passeridae: Passeriformes
Table II. – List of species currently included in Dermanyssus and their known host species, established with the help of following refe- rences: Berlese & Trouessart (1889), Bory de Saint-Vincent (1828), Dusbabek & Cerny (1971), Evans & Till (1962), Evans & Till (1964), Ewing (1922, 1933), Fain (1993), Fend’a & Schniererová (2004), De Geer (1778), Gu & Ting (1992), Haitlinger (1987), Hermann (1804), Krantz (1959), Moss (1966), Moss (1978), Moss (1970), Nelson & Furman (1967), Nosek & Lichard (1962), Schrank (1776), Uchikawa & Kitaoka (1981), Uchikawa & Takahashi (1985), Vitzthum (1921), Zeman (1979), Zeman & Jurík (1981), Zemskaya (1971), Fend’a (2006, unpublished data) and collection data from Pr. A. Fain. Taxonomic bird data follow Peterson’website. with chelae strongly reduced (Fig. 2); cornicules mem- SPECIES SPECIFIC CHARACTERS branous, flexible (not acute as in free-living mesostig- mate species) and convergent; Podosoma-legs: coxae WITHIN THE GENUS without spurs. . Adult males differ from adult females mainly in TRADITIONAL SYSTEMATICS having more extensive sclerotization both ventrally and dorsally (holoventral shield in most cases, larger dorsal ere will be dealt with characters which have shield, including more dorsal setae than in female) and been used as arguments for species description, modified chelicerae (less elongated and much broader Hnot with diagnostic characters. For all the follo- than in female, chelae with a long spermadactyl on the wing characters, only adult females are to be conside- movable digit). The tarsi of legs III and IV bear a tooth- red. like protuberance. Moreover, the genital orifice is cons- Most of specific-level discriminant characters are based picuous and presternally situated. on chaetotaxy of the legs and dorsal shield and on the relative length of peritremes against the position from KEY FOR DERMANYSSID GENERA terminating over coxae IV to over coxae III-I. Few other morphological characters are used. A marked Few characters remain currently available for dia- difference of the dorsal setae length is very conspi- gnosis between Dermanyssus and Liponyssoides. Hirst cuous between central setae on dorsal shield (j4-j6 + (1913) described Liponyssoides as a subgenus of Der- “J” series except J5) and the other setae, which are manyssus mainly based on a weak difference in the situated all around (J5, “z-Z” series, “r-R” series, “s” proportion of capitulum. Moss (1967) stated other series), in the seven following species: D. alaudae, clear differences between them based on sternal shield D. brevis, D. brevirivulus, D. hirsutus, D. grochovskae, shape and chaetotaxy of dorsal shield, sternal shield D. quintus, D. rwandae. In these species, the central and legs. But the three species L. intermedius, D. tro- seta length is near one-quarter the length of the per- chilinis and partially D. antillarum appear as yet ipheral ones, whereas they are all subequal in the other intermediate between both genera with chaetotaxy. As Dermanyssus species. a result, only following elements can be used for dia- A character concerning dorsal shield development is gnosis of the two genera. found only in five species: mesonotal scutella are pre- Sternal shield roughly crescent-shaped; usually para- sent only in D. americanus, D. antillarum, D. trans- sitic on birds……………………Dermanyssus vaalensis, D. triscutatus and D. wutaiensis. The nature Sternal shield roughly hexagonal; parasitic on of these platelets seems to correspond to the primary rodents...... Liponyssoides dorsal plates, which don’t become coalesced as they
Parasite, 2007, 14, 87-100 Mise au point 95 ROY L. & CHAUVE C.M. HISTORICAL REVIEW OF THE GENUS DERMANYSSUS
do in other Dermanyssus species in the adult stage. The posterior margin or to middle of coxa I” in D. trochi- one from 1968 and took three new species into account cient to morphologically identify this species. As a result dorsal shield is more strongly reduced in D. triscutatus linis and extends “anterad of base of coxa I” in D. faral- (D. antillarum, D. trochilinis and D. faralloni). It pro- D. hominis should be synonymized with D. gallinae. and D. antillarum than in the three other species: from loni. Although the character states seem to overlap each vided a variety of information on the genus Derma- Two species are considered incertae sedis: D. passe- the “J” series, only J1 is on the shield in D. triscutatus other, it takes part in several new species arguments. nyssus (hosts, phylogenetic results from Moss (1967), rinus Berlese & Trouessart, 1889 and D. longipes Ber- and in D. antillarum, whereas at least J1 and J2 are on For instance, the first of the two characters which are remarks on the high variability of some characters,...). lese & Trouessart, 1889. Evans & Till (1962) considered the shield in D. americanus, D. transvaalensis and cited by Uchikawa as a distinctive property in D. nip- The status of the two subgenera and the two species- them incertae sedis because of the inadequate des- D. wutaiensis. Moreover, the dorsal scutella are pro- ponensis is the remarkable length of the peritreme. Its groups is re-examined with the changes induced by the criptions and the damaged type material in the Ber- portionally quite smaller in these three latter species length is also the first of the five characteristics used three new species. Actually, the introduction of lese collection. In Moss (1978), a part of the discus- than in the two former. by Nelson and Furman in order to distinguish D. faral- D. antillarum and D. trochilinis into Moss’ phenetic sion refers to both of these species, in order to begin Three very conspicuous characters are found each in loni from the a priori most closely related species analysis introduced some problems with the subdivi- to solve this problem. In short, Zemskaya suggested a single species and concern particular opisthoventral D. hirundinis. Moreover, Oudemans (1939) cited this sions. Nevertheless, Moss decided to continue to reco- that D. passerinus should be conspecific with D. ame- setae: character as the main difference distinguishing D. che- gnize the two subgenera and the two species-groups ricanus (which should be then a junior synonym), but - a ventral neotrichy in form of a cluster of elongate, lidonis from D. hirundinis and D. gallinae. It would be temporarily because he was expecting some additional did not demonstrate this. Moss adds an argument: simple setae laterad of the anal shield is present only of interest to investigate the reliability of such a cha- elements from a new study he was preparing at that both hosts of these species seem to be conspecific too, in D. hirsutus; racter as an argument for species description. time. Unfortunately, this study was never published. according to an ornithologist. Moreover, according to - a U-shaped row of very large and deeply rooted setae In short, many of the main traditionally used species As well as these doubts about the interrelationships of Moss, it is most likely that D. passerinus and D. lon- on the opisthogaster is present only in D. quintus; specific characters are problematic. Chaetotactic cha- species in the genus Dermanyssus, Moss warned in gipes are conspecific. The type of D. longipes is too - several distally inflated setae situated posteriorly on racters of legs and dorsal shield are variable intraspe- 1978 that any person who wanted to attempt identifi- opaque to confirm such a hypothesis. the idiosoma are present only in D. antillarum. cifically and the peritreme relative length doesn’t seem cation with his key to remain careful because of the However, as the type of D. longipes (No. 52-47) is No other Dermanyssus species possess such species spe- to provide states of characters that are precise enough. considerable variation within species: “The most useful almost opaque and essentially unusable according to cific, apomorphic characters. setae for identification are those in the “j” series of the Moss and as this species has not been cited for a long Additionally, in D. quintus, the wider than long anal AN ATTEMPT WITH NUMERIC TOOLS dorsum. [...] Leg setae are also useful in species iden- time, it could be more appropriate to establish it as a plate is another species specific character. tification, but tend to be more variable within species nomen dubium instead of incertae sedis. Indeed, the Nevertheless, apart from these characters, it should be Moss tried to use more objective tools than the tradi- than dorsal setae. [...] Variation within species is consi- systematic position within Dermanyssus does not noted that the leg and dorsal shield chaetotaxy provides tional systematics in order to explore the relationships derable. One is advised to consider several characters appear to be doubtful, compared to its precise iden- most of the characters traditionally used for species dis- between the Dermanyssus species: in 1967, he publi- from several individuals of a sample when attempting tity, which is doubtful. crimination in the genus Dermanyssus. The reliability of shed a work on some numeric taxonomy theories an identification. A recent example of such variation As for D. passerinus, the type specimens are partially some of them seems to be doubtful. Evans & Till (1962) (phenetics), in which he used the genus Dermanyssus is described below for D. gallinoides. Most key cou- opaque and some papers include it in some acaro- emphasized with several remarks many intraspecific as a model. He complained about the difficulties of fin- plets list several alternative features in view of this pro- faunal lists (Nosek & Lichard, 1962; Zemskaya, 1971; variations, concerning the chaetotaxy: “The chaetotaxy ding characters in this genus and considered this to blem.” (Moss, 1978:633-634). Moss also invalidated Zeman & Jurík, 1981). So, the problem is more impor- of the venter of the opisthosoma shows considerable be due to a reduction and loss of structures correlated one of the discriminant characters for D. gallinoides: tant with this species, which we suggest be considered intraspecific variation” (p. 277). “The chaetotaxy of the with parasitism. As a result, he selected quantitative cha- “Dorsal shield scaling has been a reliable feature until species inquirenda. various segments [of the legs] is considerably more racters, which are much easier to find than qualitative recently for the separation of D. gallinoides and D. gal- D. lacertarum and D. richiardii are also problematic variable, both inter- and intra-specifically, than in the ones. His analysis resulted in two subdivisions of the linae. Two specimens just provided by N. Wilson have species. D. richiardii had been collected on two dif- free living and facultative parasitic Laelaptidae” (p. 278). genus: two subgenera, Dermanyssus and Microder- predominantly scaled teeth in one case and comple- ferent species of insect, the hymenopteran Xylocopa Moss (1968) also noticed that characters of the leg manyssus, are to be distinguished from one another tely scaled teeth in the other, but otherwise key to violaceus and the lepidopteran Cossus ligniperda (Canes- chaetotaxy usually seemed to be the most variable. by the setation of genu II (two al setae, one av and D. gallinoides” (Moss, 1978:634) trini & Fanzago, 1877), which are not common hosts Other than the chaetotactic characters of the legs and one pv setae in Dermanyssus (Dermanyssus); one al for the Dermanyssus. As for D. lacertarum, it was trans- setae, no av and pv setae in Dermanyssus (Microder- dorsal shield, the other traditionally most used charac- SPECIES WHOSE NOMENCLATURAL STATUS IS NOT CLEAR ferred from genus Ricinus by Canestrini (1877) in the ter is the relative length of the peritreme. However the manyssus)) and by the size of the unengorged female’s same paper, with the only following sentence: “Due states of this character don’t provide any defined limit. body. Subgenus Dermanyssus included D. chelidonis, About D. hominis, Dugès wrote as follows: “D. homi- specie vi sono citate come nuove, il Ricinus lacer- The extension of the peritreme varies continuously D. gallinae, D. gallinoides, D. grochovskae, D. hirsutus, nis; sorte d’acaride, Bory St-Vincent”. Bory de Saint-Vin- tarum, e l’Acarus penetrans; il primo sembra un Der- from coxa IV to coxa III to coxa IV to coxa I without D. hirundinis, D. prognephilus, D. quintus, D. trans- cent described it in 1823, in a memoir which was read manyssus, il secondo è una forma larvale”. From that a sharp gap from one species to another in the genus vaalensis and D. triscutatus. Subgenus Microdermanys- during a regular meeting of the French “Académie des date, we did not find any more information on these P
Dermanyssus and with intraspecifical variations. Indeed, sus included D. alaudae, D. americanus and D. brevis. Sciences”. He did not attribute any name and any sys- two species. Maybe they should be established as UBLICATION from Moss (1978) the following data can be extracted, Moreover, two species-groups were separated in the tematic position within the Acari group to this species. nomina dubia. In any case, given the information in increasing order: the peritreme extends from the subgenus Dermanyssus: the hirsutus-group, with, Latreille and Savigny were designated to judge such a cited above, they can not be included in the genus coxa IV “to middle of coxa III” in D. transvaalensis among other discriminant characters, the seta al1 of position, but did not do so (cf. Académie des Sciences, Dermanyssus anymore. and D. chelidonis, “not as far as anterior margin of coxa palp genu spiniform, and the gallinae-group, with the 1823-1828). Mites of this species were found infesting III” in D. alaudae and D. brevis, “to or past anterior seta al1 of palp genu spatulate. The hirsutus-group the body of a woman. The description of this species margin of coxa III” in D. americanus, “past anterior included D. grochovskae, D. hirsutus and D. quintus, and its illustration by Bory de Saint-Vincent (1828) CONCLUSION AND PERSPECTIVES margin of coxa III” in D. triscutatus, “to middle of whereas the gallinae-group included D. chelidonis, appear very similar to D. gallinae. This last species has coxa II” in D. hirsutus, “to middle or anterior margin D. gallinae, D. gallinoides, D. hirundinis, D. progne- been reported from humans several times (Beck, 1999; he systematics of Dermanyssus is not completely of coxa II” in D. gallinae and D. gallinoides, “to ante- philus, D. transvaalensis and D. triscutatus. Moss used Cremer & Morrien, 1962; Holz J., 1954; Pampiglione et clear as yet. Its history is complex. Dermanyssus rior margin of coxa II” in D. grochovskae and D. hirun- these new subdivisions in his 1968 and 1978 keys. The al., 2001). Moreover, there is no type material available Tseems to be well defined today, but species dinis, “to middle of coxa I” in D. prognephilus, “past 1978 article was somewhat more developed than the and the drawing of Bory de Saint-Vincent is not suffi- within the genus remain less clearly defined. Moreover, I Parasite, 2007, 14, 87-100 Parasite, 2007, 14, 87-100 96 Mise au point Mise au point 97 PUBLICATION I
ROY L. & CHAUVE C.M.
not only was the last work on Dermanyssus not com- BORY DE SAINT-VINCENT J.-B.G.M. Note sur une espèce d’Aca- pletely carried out, but also five new species have been ride qui vit sur le corps humain. FERUSSAC, Bulletin des described since this last review (Moss, 1978). The relia- Sciences Naturelles, 1824, II, 305-306. bility of numerous traditional characters needs to be BORY DE SAINT-VINCENT J.-B.G.M. Sur un nouveau genre re-examined. d’Ascaridiens sorti du corps d’une femme. Audouin et al. Today, 23 species are included in this genus, two of Annales des Sciences Naturelles, 1828, XV, 125-131. which are really doubtful. The status of D. longipes and CANESTRINI & FANZAGO INTORNO AGLI ACARI ITALIANI. Atti del Regio D. passerinus is to be re-examined. D. americanus Istituto veneto di scienze lettere ed arti, 1877, IV (5:1-4), might be a junior synonym of D. passerinus. D. homi- 57-58. nis is a synonym of likely D. gallinae. Some other early CONTARINI N. Cataloghi degli uccelli e degli insetti delle pro- species might likely be synonymized too. The reliabi- vincie di Padova e Venezia, Ed. Bassano, 1843. lity of the five species described after Moss (1978), CREMER G. & MORRIEN J.J. Dermanyssus scabies in man. Ned needs to be checked and they have to be integrated Tijdschr Geneeskd., 1962, 17 (106), 520-523. in a review of the entire genus. DUGÈS A.L. Recherches sur l’ordre des Acariens en général For these reasons, it appears necessary to review the et la famille des Trombidiés en particulier. Annales des genus Dermanyssus at the specific level, which we plan sciences naturelles. Zoologie, 1834, 2 (1), 5-46, pl. 1. to do, with the help of cladistic tools. Two major ques- DUSBABEK F.& CERNY V. Two mesostigmatic mites (Acarina: tions need to be answered. First, the correct definition Macronyssidae and Dermanyssidae) associated with Cuban of genus, even if it seems to be right using traditional birds. Folia parasitologica, 1971, 18, 55-61. tools, has to be checked by testing the monophyly of EVANS G.O. Observations on the chaetotaxy of the legs in the group. Secondly, the a priori most problematic the free-living Gamasina (Acari: Mesostigmata) Bulletin of question of the species definitions within the genus the British Museum (Natural History) Zoology, 1963, 10, 277-303. should be explored, and maybe some species should be synonymized. Finally, as morphological characters EVANS G.O. Observations on the ontogenetic development of seem to be insufficient, it seems necessary to add mole- the chaetotaxy of the tarsi of legs II-IV in the Mesostig- mata (Acari). Proceedings of the II International Congress cular characters to the phylogenetic analysis. of Acarology (1967), 1969, 195-200. EVANS G.O. & TILL W.M. The Genus Dermanyssus De Geer (Acari: Mesostigmata). Annals and magazine of Natural ACKNOWLEDGEMENTS History, 1962, 13 (5), 273-293. e are grateful for the help and advice of EVANS G.O. & TILL W.M. A new species of Dermanyssus and a redescription of Steatonyssus superans Zemskaya (Acari: P. Fend’a (Faculty of Natural Sciences, Come- Mesostigmata). Acarologia, 1964, 6 (4), 624-631. nius University, Bratislava, Slovak Republic), W EVANS G.O. & TILL W.M. Studies on the British Dermanyssi- R. Nannelli (Istituto Sperimentale per la Zoologia Agra- dae (Acari: Mesostigmata). Part I. External morphology. ria, Firenze, Italia), J. Hallan (Texas A & M University, Bulletin of the British Museum (Natural History) Zoology, College Station, Texas), O. Bain and I. Roy (Muséum 1965, 13, 249-294. National d’Histoire Naturelle, Paris, France), H. Ferrière EVANS G.O. & TILL W.M. Studies on the British Dermanys- (Université Paris I-Sorbonne, Paris, France), V. Marengo sidae (Acari: Mesostigmata). Part II. Classification. Bulletin (Muséum d’Histoire naturelle, Lyon, France) and to Pr. of the British Museum (Natural History) Zoology, 1966, 14, F.J. Radovsky and an anonymous reviewer for cons- 109-370. tructive reviewing and valuable suggestions. EWING H.E. Dermanyssus americanus, new species. Procee- dings of the United States National Museum, 1922, 62 (13), 24-26, pl. 2. REFERENCES EWING H.E. The dermanyssid mites of North America. Pro- ceedings of the United States National Museum, 1923, 62, ACADÉMIE DES SCIENCES. Comptes-rendus hebdomadaires des 1-26, pls 1, 2. séances de l’Académie des sciences, Paris, 1820-1835, T7- EWING H.E. Dermanyssus prognephilus, new species. Pro- T10. ceedings of the United States National Museum, 1933, 82 BECK W. Farm animals as disease vectors of parasitic epi- (30), 12. zoonoses and zoophilic dermatophytes and their impor- EWING H.E. A short synopsis of the North America species tance in dermatology. Hautarzt, 1999, 50 (9), 621-628. of the mite genus Dermanyssus. Proceedings of the Ento- BERLESE A. Acari, myriopoda et scorpiones hucusque in Italia mological Society of Washington, 1936, 38, 47-54. reperta. Patavii - F. Salmin., 1889, fasc. 53. FAIN A. A new species of the genus Dermanyssus De Geer, BERLESE A. & TROUESSART E. Diagnoses d’acariens nouveaux 1778 (Acari: Dermanyssidae) from the nest of a bird Apus ou peu connus. Bulletin de la Bibliothèque Scientifique de affinis in Rwanda. Bulletin & Annales de la Société royale l’Ouest 2e année, 1889, 2e partie (9), 121-143. d’entomologie de Belgique, 1993, 129, 163-168.
Parasite, 2007, 14, 87-100 98 Mise au point PUBLICATION I
HISTORICAL REVIEW OF THE GENUS DERMANYSSUS
FEND’A P. & SCHNIEREROVÁ E. Mites (Acarina: Mesostigmata) to the species (Acari: Mesostigmata: Dermanyssidae). Jour- in the nest of Acrocephalus spp. and in neighbouring nal of Medical Entomology, 1978, 14 (6), 627-640. reeds. Biologia, Bratislava, 2004, 59 (Suppl. 15), 41-47. MOSS W.W. & RADOVSKY F.J. Dermanyssus hirsutus, a new GEER CH. de. Mémoires pour servir à l’histoire des insectes, name for Dermanyssus scutatus Ewing (Acari: Mesostig- 1778, VII, 111-112. mata: Dermanyssidae). Journal of the Kansas Entomological GU Y.M. & TING Q.Y. New species and new records of Der- Society, 1967, 40, 277. manyssus from China (Acari: Dermanyssidae) Acta Zoo- MOSS W.W., MITCHELL C.J. & JOHNSTON D.E. New north ame- taxonomica Sinica, 1992, 17 (1), 32-36 (abstract in English, rican host and distribution records for the mite genus Der- text in Chinese). manyssus (Acari: Mesostigmata: Dermanyssidae). Journal HAITLINGER R. Dermanyssus alaudae (Schrank, 1781) and of Medical Entomology, 1970, 7, 589-593. other mites (Acari: Dermanissidae, Macronyssidae, Hae- NELSON B.C. & FURMAN D.P. A new species of Dermanyssus mogamasiade, Hirstionyssidae, trombiculidae, Erythraeidae) from marine birds, with observations on its biology (Aca- collected from birds in Poland. Wiadomos Parazytolo- rina: Dermanyssidae). Acarologia, 1967, 9, 330-337. giczne, 1987, 33 (2), 233-245 (in Polish). NOSEK J. & LICHARD M. Beitrag zur kenntnis der vogelnest- HERMANN J. Mémoire aptérologique. Levrault (ed.), Stras- fauna. Entomologické problémy (Bratislava), 1962, 2, 29-51. bourg, 1804. OUDEMANS A.C. Kritisch historisch overzicht der acarologie. HIRST S. On three new species of Gamasid mites found on Tweede gedeelte, 1759-1804. Tijdschrift voor entomologie, rats. Bulletin of entomological research, 1913, 4, 119-124. 1929, 72 (suppl.), 99. OLZ H J. Dermanyssus avium an occasional parasite in man. OUDEMANS A.C. Kritisch historisch overzicht der acarologie. Arztl Wochensch. Jan, 1954, 2, 9 (1), 16. Derde gedeelte, 1805-1850. Vols A.-G. Leiden, E.J. Brill, HUGHES T.E. Mites or the Acari. Athlone Press, University of 1936-1937. London, 1959. OUDEMANS A.C. Neue funde auf dem gebiete der systematik ICZN. International code of zoological nomenclature, 4th ed, und der nomenklatur der Acari, V. Zoologische Anzeiger, London, 1999. 1939, 126 (11-12), 303-309. KOCH C.L. Deutschlands crustaceen, myriapoden und arach- PAMPIGLIONE S., PAMPIGLIONE G., PAGANI M., RIVASI F. Persis- niden. Ein Beitrag zur deutschen Fauna, Dr. Herrich- tent scalp infestation by Dermanyssus gallinae in an Emi- Schäffer, 1836, 4, 14. lian country-woman. Parassitologia, 2001, 43 (3), 113-115. KOLENATI F.A. Beiträge zur kenntniss des arachniden. Sit- PETERSON A.P. http://www.zoonomen.net/avtax/frame.html, zungsberichte der Kaiserlichen Akademie der Wissenschaf- Version 7.07 (2007.02.04). ten, 1859, 35, 155-190, 8 tab. PHILLIS III W.A. Ultrastructure of the chelicerae of Derma- KRANTZ G.W. New synonymy in the Dermanyssidae Kolenati, nyssus prognephilus Ewing (Acari: Dermanyssidae). Inter- 1859, with a description of a new species of Dermanyssus national Journal of Acarology, 2006, 32 (1), 85-91. (Acarina, Dermanyssidae). Proceedings of the Entomolo- gical Society of Washington, 1959, 61, 174-178. RADOVSKY F.J. Revision of the Macronyssid and Laelapid mites of bats: outline of classification with descriptions of LINDQUIST E.E. & EVANS G.O. Taxonomic concepts in the new genera and new type species. Journal of Medical Ascidae, with a modified setal nomenclature for the idio- Entomology, 1966, 3(1), 93-99. soma of the Gamasina (Acarina: Mesostigmata). Memoirs of the Entomological Society of Canada, 1965, 47, 1-64. RADOVSKY F.J. The Macronyssidae and Laelapidae (Acarina: Mesostigmata) parasitic on bats. University of California MELVILLE R.V. & SMITH J.D.D. Official lists and indexes of Publications in Entomology, 1967, 46, 1-288. names and works in zoology. International Trust for Zoo- logical Nomenclature, London, 1987. RADOVSKY F.J. Adaptative radiation in the parasitic Mesostig- mata. Acarologia, 1969, 11, 450-483. MOSS W.W. Dermanyssus gallinoides n. sp. (Mesostigmata: Laelapoidea: Dermanyssidae), an acarine parasite of wood- SHEALS G.J. The status of the genera Dermanyssus, Alloder- th peckers in Western North America. The Canadian Ento- manyssus and Liponyssoides. Proceeding of the XI Inter- mologist, 1966, 98, 635-638. national Congress of Entomology, Vienna, 1962, 2, 473- 476. MOSS W.W. The biological and systematic relationships of the martin mite, Dermanyssus prognephilus Ewing (Acari: Meso- SCHRANK F. von P. Beyträge zur naturgeschichte. Siebente Art, tigmata: Dermanyssidae). Ph. D. dissertation, The Univer- Augsburg, 1776, p. 11. sity of Kansas, Lawrence, 1966. SCHRANK F. von P. Enumeratio insectorum austriae indigeno- MOSS W.W. Some new analytic and graphic approaches to rum. Augustae Vindelicorum: vid. Eberhardi Klett et Frank, numerical taxonomy, with an example from the Derma- 1781. nyssidae (Acari). Systematic zoology, 1967, 16, 177-207. STRANDTMANN R.W. & WHARTON G.W. A manual of mesostig- MOSS W.W. An illustrated key to the species of the Acarine matid mites parasitic on vertebrates. Institute of Acarology, genus Dermanyssus (Mesostigmata: Laelapoidea: Derma- Contribution No 4. University of Maryland, College Park, nyssidae). Journal of Medical Entomology, 1968, (1), 67- Maryland, 1958. 84. TENORIO J.M. Catalog of the world Echinonyssus (= Hirstio- MOSS W.W. The mite genus Dermanyssus: a survey, with des- nyssus) (Acari: Laelapidae). International journal of ento- cription of Dermanyssus trochilinis n. sp., and a revised key mology, 1984, 26 (3), 260-281.
Parasite, 2007, 14, 87-100 Mise au point 99 ROY L. & CHAUVE C.M.
UCHIKAWA K. & KITAOKA S. Dermanyssus nipponensis sp. nov. taken from Japanese green woodpecker indigenous to LETTRE À LA RÉDACTION Japan (Acari: Mesostigmata) National Institute of Animal Health Quarterly, 1981, 21 (2), 80-82. UCHIKAWA K. & TAKAHASHI M. Contribution to mites of the FONDATI A. Efficacy of daily oral ivermectin in the treatment genus Dermanyssus De Geer distributed in Japan (Acarina, of 10 cases of generalized demodicosis in adult dogs. Vete- Mesostigmata) Japanese Journal of Sanitary Zoology, 1985, rinary Dermatology, 1996, 7, 99-104. 36 (3), 233-237. HILLIER A. & DESH C.E. Large-bodied Demodex mite infesta- VITZTHUM H.G. Acarologische Beobachtungen. 4 Riehe. Archiv tion in four dogs. Journal of American Veterinary Medical für Naturgeschichte, 1921, 86A (10), 1-69. Association 2002, 220, 623-627. VOLS A.G. LEIDEN E.J. Brill, 1936-1937. MEDLEAU L., RISTIC Z. & MCCELVEEN D.R. Daily ivermectin for ZEMAN P. Dermanyssus carpathicus sp. n. (Acarina: Derma- treatment of generalized demodicosis in dogs. Veterinary nyssidae), a new bird parasite from Czechoslovakia. Folia Dermatology, 1996, 7, 209-212. Parasitologica (Praha), 1979, 26, 173-178. MOZOS E., PEREZ J., DAY M.J., LUCENA R. & GINEL P.J. Leish- ZEMAN P. & JURÍK M. A contribution to the knowledge of fauna maniosis and generalized demodicosis in three dogs: a cli- and ecology of gamasoid mites in cavity nests of birds in nicopathological and immunohistochemical study. Journal Czechoslovakia. Folia Parasitologica (Praha), 1981, 28, of Comparative Pathology 1999, 120, 257-268. 265-271. PURVIS A.C. Immunodepression in Babesia microti infections. ZEMSKAYA A.A. A New species of bird mites, Dermanyssus gro- Parasitology, 1977, 75, 197-205. chovskae Zemskaya sp. n. Zoologiceskij Zurnal, 1961, 40, SARIDOMICHELAKIS M., KOUTINAS A., PAPADOGIANNAKIS E., PAPA- 134-136. ZACHARIADOU M., LIAPI M. & TRAKAS D. Adult-onset demo- ZEMSKAYA A.A. Mites of the family Dermanyssidae Kolenati, dicosis in two dogs due to Demodex canis and a short- 1859, of the USSR fauna. Medicinskaâ Parazitologiâ i tailed demodectic mite. Journal of Small Animal Practice, Parazitarnye bolezni, 1971, 40, 709-717 (in Russian). 1999, 40, 529-532. Reçu le 15 novembre 2006 TARELLO W. Cutaneous lesions in dogs with Dirofilaria (Noch- Accepté le 12 mars 2007 tiella) repens infestation and concurrent tick-borne trans- mitted diseases. Veterinary Dermatology, 2002, 13, 267- 274. TARELLO W. Concurrent cutaneous lesions in dogs with Babesia gibsoni infection in Italy. Revue de Médecine Vété- rinaire, 2003a, 154, 281-287. TARELLO W. Canine granulocytic ehrlichiosis (CGE) in Italy. Acta Veterinaria Hungarica, 2003b, 51, 73-90.
Reçu le 15 juin 2007 Accepté le 20 septembre 2007 P UBLICATION
Erratum
ROY L. & CHAUVE C.M. Historical review of the genus Dermanyssus Dugès, 1834 (Acari: Mesostigmata: Der- manyssidae). Parasite, 2007, 14 (2), 87-100. The authors wish to correct an error in the published manuscript. On page 97, column 2, line 16, a confu- sion between two species names (D. passerinus and D. hirundinis) has been found. The sentence should read: “Moreover, according to Moss, it is most likely that D. longipes and D. hirundinis are conspecific”. Authors apologize to readers for this mistake. I Parasite, 2007, 14, 87-100 Parasite, 2007, 14, 339-341 100 Mise au point Parasite, 2007, 14 341
4.2 Evaluation des caractères morphologiques discriminants entre espèces : publication II
a - Présentation
La déconcertante variabilité des caractères morphologiques disponibles pour l’appréhension du genre Dermanyssus au niveau spécifique a été plusieurs fois soulignée. Les analyses pourtant très approfondies de Moss (1967, 1968, 1978) ont au moins partiellement échoué dans la clarification des délimitations interspécifiques.
4.2.a.1 Objectifs Les objectifs consistaient ici à vérifier expérimentalement les assertions antérieures concernant l’extrême variabilité intraspécifique des caractères utilisés jusqu’alors et évaluer l’utilité de chacun en vue d’une exploration phylogénétique.
4.2.a.2 Matériel et méthodes Pour ce faire, les caractères traditionnellement utilisés pour la discrimination interspécifique dans le genre Dermanyssus ont été observés, notés, dessinés chez différents individus appartenant à quelques populations de D. gallinae. Des informations complémentaires ont été tirées de types et paratypes d’autres espèces du genre, réparties dans les trois divisions de Moss.
4.2.a.3 Principaux résultats Les principaux caractères morphologiques utilisés comme caractères diagnostiques chez les Mesostigmata, et en particulier chez les Dermanyssina, se situent dans la chaetotaxie des pattes, celle de la plaque dorsale et d’autres parties de l’idiosome et la longueur relative des péritrèmes par rapport à la position des coxae. D’une manière générale, de nombreux auteurs ont remarqué que la plupart des caractères étaient, pour des raisons non évidentes, beaucoup moins stables chez les acariens parasites que chez leurs parents libres dans ce groupe. En particulier, une étude a mis en évidence la variabilité accrue chez les Mesostigmata parasites de la chaetotaxie des pédipalpes (Evans 1963) par rapport aux Mesostigmata libres. Ainsi l'analyse de la chaetotaxie des pattes chez les acariens (cf. Fig. 5) repose-t-elle sur le nombre et la position de soies sur des portions définies dans la surface de chacun des segments, de forme plus ou moins cylindrique ou conique (ad, antéro-latérale, pv postéro-ventrale, …). Or, très souvent, l’aire d’implantation de soies homologues chez Dermanyssus est apparue très élargie par comparaison aux Mesostigmata libres, et s’est montrée chevauchante avec les aires limitrophes. De fait, il s’avére souvent très difficile de décider à laquelle des portions appartiennent les soies observées. Sans même parler des soies absentes ou supplémentaires au sein d’une population de la même espèce, voire au sein d’un seul individu.
53 Patte I Patte II
Tarse
Tibia
Patte III Patte IV Genou av ad al Fémur
Figure 5. Cheatotaxie des pattes chez les acariens mésostigmates. A gauche, notation de la chaetotaxie des pattes chez les acariens mésostigmates d’après Evans (1963). Exemple de la représentation diagrammatique de la chaetotaxie du genou des pattes I à IV chez les deutonymphes et adultes de Pergamasus.a=antéro- p=postéro-, d=dorsal, v=ventral, ex. zone ad= zone antéro- dorsale. A droite, représentation en 3 dimensions d’une patte stylisée d’acarien mésostigmate avec surlignage coloré des différentes zones à observer. Les limites de ces zones sont virtuelles et s’avèrent extrêmement floues dans le genre Dermanyssus, les bases des soies trouvant souvent leur implantation à la limite de la zone, voire à l’extérieur, empiétant sur la zone voisine, de manière asymétrique sur un seul et même individu.
L’article suivant démontre que la variabilité des caractères traditionnels, jusqu’alors seulement évoqués à propos de Dermanyssus (Moss 1978), est extrême et rend leur utilisation dans le cadre d’une reconstruction phylogénétique très délicate. La mise en évidence de chevauchement de nombreux états entre espèces dans le genre, de variations importantes au sein d’une unique population de D. gallinae, ainsi que la récurrence d’asymétries bilatérales sont des arguments de poids en faveur de la recherche de caractères « nouveaux », non décrits jusqu’alors, ou peu utilisés.
b - Remarques sur la publication II
4.2.b.1 Des caractères réhabilités a posteriori Parmi les caractères dénoncés comme douteux dans la publication II, plusieurs ont été testés au cours de la phase a posteriori de la publication III (p. 456-457 et 461). Pour une majorité des caractères testés concernant la chaetotaxie des pattes en particulier, l'instabilité est confirmée. Toutefois, certains en ressortent réhabilités.
54 Publication II
c - Publication II Proceedings of the XIth International Congress of Acarology, Amsterdam, August 2006
The genus Dermanyssus Dugès, 1834 (Acari : Mesostigmata : Dermanyssidae): species definition
Running title: Species definition within Dermanyssus
L. Roy* and C. Chauve Laboratoire de Parasitologie et Maladies Parasitaires, Ecole Nationale Vétérinaire de Lyon 1 avenue Bourgelat, 69280 MARCY-L'ETOILE, France, Email: [email protected], Tel 00 33 4 78 87 25 74, Fax 00 33 4 78 87 25 77
Abstract The genus Dermanyssus Dugès, 1834 (Acari: Mesostigmata: Dermanyssidae) includes species of hematophagous mites that are ectoparasites of birds. The definition of this genus took a long time to get firmly established. Major changes in genus definition involving not only the establishment of synonymies, but also many changes in systematic position have been reviewed, based on literature data, by Roy and Chauve (2007). However, the species definition is currently not yet clear. Host specificity and geographic distribution of Dermanyssus species are reviewed. Some morphological characters posing problems in species identification are discussed.
Key-words: Acari, Mesostigmata, Dermanyssus, morphological characters, systematics
Introduction The genus Dermanyssus Dugès, 1834 (Acari: Mesostigmata: Dermanyssidae) includes hematophagous mite species that are ectoparasites of birds. One of the species in this genus, D. gallinae (De Geer, 1778), the Red Fowl Mite, is economically important in the poultry industry. The damage consists of downgraded eggs, decreased egg production, anaemia and even mortality due to exsanguination. D. gallinae can also transmit diseases, such as avian spirochaetosis, fowl cholera and salmonellosis (Valiente Moro et al. 2005). Although the genus harbours species of economic importance, the classification of species in this genus has been in a state of confusion. The main characters discriminating Dermanyssidae from other families in the Dermanyssoidea are located on the chelicerae. Dermanyssidae have a strongly elongated 2nd segment of the chelicerae and they have much reduced chelae, a morphological feature that seems to be correlated to a hematophagous life style (see Phillis 2006, for more details). Macronyssidae, another family in the Dermanyssoidea, also includes obligatory hematophagous species which possess chelicerae modified but in a different way. Here, elongation is less important and concerns the first article, rather than the second as in Dermanyssidae and the chelae are not really atrophied as in the Dermanyssidae. The current Family Dermanyssidae includes only two genera: Dermanyssus and Liponyssoides. The main morphological differences between those two genera are to be found on genu IV (ad with 2 setae in Dermanyssus / 3 setae in Liponyssoides) and on the sternal shield (roughly crescent-shaped with 1 or 2 pairs of setae in Dermanyssus / roughly hexagonal with 3 pairs of setae in Liponyssoides). The host range differs
55 Publication II widely between the two genera. Whereas Dermanyssus includes bird parasites, Liponyssoides includes species mainly parasitic on rodents. Only Evans and Till (1962) and Moss (1967, 1968, 1978) provided comprehensive systematic reviews of genus Dermanyssus at the species level. Although the genus seems to be well defined today, this does not hold for the species within this genus (Roy and Chauve 2007). At least 57 species have been included in Dermanyssus, whereas currently there are 23 species (Roy and Chauve 2007). Here, we provide an overview of the current characterisation of species within the genus Dermanyssus. Firstly, we review the host specificity and geographic distribution of Dermanyssus species, based on the data published to date. Secondly, we examine the reliability of traditional species-specific characters.
Geographic distribution of Dermanyssus species Concerning the geographical distribution of Dermanyssus species, D. gallinae is the most frequently collected species and it seems to be found all around the world. Some other Dermanyssus species are also cosmopolitan (see distribution map in Roy and Chauve 2007). Examples are D. hirundinis, D. brevis and D. quintus which have been reported from both American and Eurasian continents and are thus not restricted to the New or Old World. Some other Dermanyssus species stem from a single and recent record, such as D. antillarum Dusbabek & Cerny, 1971 (Cuba), D. nipponensis Uchikawa & Kitaoka, 1981 (Japan) and D. rwandae Fain, 1993 (Rwanda). Clearly, there are not enough data on these species to make any inference on their distribution.
Host specificity Dermanyssus species are ectoparasites of birds. However, most species do not show host specificity, and some species have even been noted parasitizing mammals, such as man or rodents, in absence of birds as hosts. Up to 30 bird species, distributed in twelve different families of birds and in eight different orders, are known to be parasitized by D. gallinae (Roy and Chauve 2007). Also, D. hirundinis has a rather broad host spectrum since it has 40 bird species as hosts out of 18 families and 9 orders of birds (bird classification from Peterson 2007). There may be a few species that are host-specific. Examples are D. alaudae, D. quintus, D. brevis and D. triscutatus each of which have been found on birds belonging to a single family. The most recently described Dermanyssus species might be host- specific too (Table 2), but there are not enough data available to prove that this is not a simple byproduct of limited sampling.
The question of the reliability of species specific characters Traditional systematics Most of the species-level discriminating characters are based on leg and dorsal chaetotaxy, and relative length of the peritreme against coxal position. Concerning leg chaetotaxy, some authors (Evans and Till 1962, Moss 1978) cautioned that there is great intraspecific variation. As for dorsal chaetotaxy and relative peritreme length, characters also seem to be very polymorphic (Roy, pers. obs.). a - Major characters: dorsal shield chaetotaxy and relative length of peritreme An example of the dorsal shield chaetotaxy from a single population of D. gallinae is shown in Figure 1. Focusing on the j line allows observing some important variation in the dorsal chaetotaxy. Dispositions of setae J3, J4 and J5 are shown (two nearly parallel longitudinal lines in c or a hexagone in f and i). Several cases of bilateral asymmetries in setation are found, not only with
56 Publication II some asymmetrical dispositions (e, j), but also with a case of unpaired seta J3/J4 (d with only one of them on the right side). Note that such frequent asymmetries have been pointed out by Evans and Till (1962). Moreover, there are major differences in the shape of the dorsal shields of mites shown in Figure 1. The shape of the dorsal shield is important for evaluating some chaetotactic characters: variation in position of j1 , i.e. on or off shield, seems to be due to shield contour variations. Arrows in Figure 1 show different positions of j1, which is always on-shield in D. gallinae, but off-shield in D. gallinoides according to Moss (1978). Allred (1970) pointed out a similar case in two species of Ornithonyssus (Mesostigmata: Macronyssidae), which are also hematophagous parasites. In these species, the great degree of intraspecific variation in shape and chaetotaxy of the sternal plate imposes major difficulties for species discrimination. The peritreme is associated with the respiratory organ. It is a groove extending anteriorly from the stigma, which is located near coxa IV. Using a scanning electron microscope, two sclerotized lips can be seen along the groove (Figure 3). The relative length of the peritreme from coxa IV to coxa III-I is considered to be a taxonomic character discriminating between Dermanyssus species, but there are numerous character states and they overlap each another (Roy, pers. obs.; data from Moss 1978). The extension of the peritreme varies continuously from coxa IV to coxa III up to coxa I without a distinct gap throughout the genus Dermanyssus as well as within species of this genus. There are more than 7 possible positions (Figure 2): peritreme extending to anterior margin of coxa III, and to posterior, middle or anterior margin of coxae II and I, and a various intermediate positions can be observed (Roy, pers. obs.). Moreover, clearing specimens for slide-mounting and observation with an optical microscope may destroy attachment of the peritreme, so that the soft groove may change length and position, moving inside the podosoma (Figure 4 illustrates it with a case of asymmetric length). However, the clearing procedure is necessary for unambiguous identification of hematophagous mites. Phenotypic plasticity concerning these frequently used characters and overlapping character states make them very difficult to be encoded for cladistic exploration.
b - Other characters Apart from the major characters dealt with previously, some other types of character have been used for species identification. Two characters concerning the dorsal side allow the characterization of two small groups of species and each seems to provide a rather distinct gap between their two states: x Dorsal setae show marked or no difference in length between "central setae" and "peripheral setae" (central setae = j4-j6 + J-serie except J5; peripheral setae = J5, z-serie, Z-serie, r-serie, R- serie, s-serie). “Central setae” are markedly shorter than "peripheral setae" in 7 species (D. alaudae, D. brevis, D. brevirivulus, D. hirsutus, D. grochovskae, D. quintus, D. rwandae). x Mesonotal scutella are present or not. They are present in 5 species (D. americanus, D. antillarum, D. transvaalensis, D. triscutatus and D. wutaiensis). Each of the following three characters defines a single species: x Ventral neotrichy in the form of a cluster of elongate, simple setae, lateral of the anal shield, is present only in D. hirsutus. x Several inflated setae situated posteriorly on the idiosoma are found only in D. antillarum. x A U-shaped row of very hard and deeply rooted setae on the opisthosomal ventral side is present only in D. quintus. In addition, several chitinous apophyses on coxae III and IV and some clawlike setae on trochanters and coxae III and IV, as well as an anal shield, more broad than long, constitute a set of unique characters for D. quintus.
57 Publication II
A new character Moss (1967), who studied the relationships between species within the genus Dermanyssus using phenetic tools, selected mostly morphometric characters in absence of clearly definable morphological characters - a feature he assumed to be an adaptation to the parasitic life style. In addition, he described a character not noted before, i.e. the shape of seta al1 of palp genu (distally expanded vs spike-like). Based on his analysis, two subgenera were distinguished, but, in 1978, the addition of three new species blurred these subdivisions. No more exploration of interrelationships between Dermanyssus species have been published thereafter.
Discussion on species definition The most frequently invoked characters are problematic: (1) leg chaetotaxy provides characters which are quite variable intraspecifically according to different authors, (2) dorsal shield chaetotaxy provides characters which are very variable at least in D. gallinae (which differs from D. hirundinis only by the number of setae pairs on dorsal shield according to Evans & Till 1962) and (3) relative length of the peritreme is not as clearly defined as required for species characterization within the genus Dermanyssus. Five additional characters seem to be reliable, but characterize only few species. Moss revealed a new not morphometric and apparently intraspecifically steady character (shape of seta al1 of palp genu). Most of the purely morphological characters are too polymorphic intraspecifically and their number is not high enough, since there are few left that are reliable enough. Species within the genus Dermanyssus are not sufficiently clearly defined so that it is hard to separate them from each based on morphological characters alone. Thus, species-specific characters, i.e. characters used to define new species, do not seem to be stable enough. Moreover, most of these species seem to provide low host specificity and are geographically widely distributed. For all these reasons there is doubt about the validity of several species within the genus Dermanyssus.
Conclusion and perspectives In conclusion, the genus Dermanyssus includes species which are morphologically not clearly defined and therefore not unambiguously distinguished from each other. This is largely due to high phenotypic plasticity in many of traditional species-specific characters. Moreover, not only the last- published review on the genus Dermanyssus was unachieved, but also five new species have been described since this last review. As a result, the genus is in need of revision, which we plan to do with the help of cladistic tools. Two major questions are to be answered: first, the correct definition of genus, even if it seems to be right with traditional tools, has to be checked by testing the monophyly of the group. Secondly, the a priori most problematic question on the species definition inside genus has to be solved. Finally, as morphological characters seem to be insufficient, it is necessary to add molecular characters to the phylogenetic analysis of the genus Dermanyssus.
Acknowledgements We are grateful for the help, advices, loans of type and non-type material, etc. of Baker A.S. and Beccaloni J. (The Natural History Museum, London, UK), Huber B. (Zoologisches Forschungsinstitut and Museum A. Koenig, Bonn, Germany), Bain O. and Judson M. (Muséum National d’Histoire Naturelle, Paris, France), Lindquist E.E., Beaulieu F. and King Wan Wu
58 Publication II
(Agriculture & Agri-Food Canada, Ottawa, Canada), Klompen H. (Ohio State University, Museum of Biological Diversity, Columbus, USA), Ochoa R. (Systematic Entomology Laboratory, USDA, Beltsville, USA), O’Connor B.M. (Museum of Zoology, University of Michigan, Ann Arbor, USA), Dowling A.P.G. (University of Kentucky, Lexington, USA), Kilpinen O. (Danish Institute of Agricultural Sciences, Lyngby, Denmark), Knee W. (University of Alberta, Edmonton, Canada), Nannelli R. (Istituto Sperimentale per la Zoologia Agraria, Firenze, Italia), Hallan J. (Texas A&M University, College Station, Texas), Marengo V. (Muséum d’Histoire Naturelle, Lyon, France).
References Allred DM (1970) – Dermanyssid mites of New Guinea. J. Med. Entomol. 7:242-246 Berlese A, Trouessart E (1889) Diagnoses d’acariens nouveaux ou peu connus. Bull Biblioth Sci Ouest 2e année, 2e partie (9) :121-143 Canestrini, Fanzago (1877) Intorno agli acari italiani. Atti R Ist veneto sci lett arti IV :5 :1-4 ;57-58 Contarini N (1843) Cataloghi degli uccelli e degli insetti delle provincie di Padova e Venezia. Tipografia Baseggio edit, Bassano Dugès AL (1834). Recherches sur l'ordre des Acariens en général et la famille des Trombidiés en particulier. Ann sci nat, Zool (2)1: 5-46 pl. 1. Dusbabek F, Cerny V (1971) Two mesostigmatic mites (Acarina: Macronyssidae and Dermanyssidae) associated with Cuban birds. Folia parasitol 18:55-61 Evans GO (1963) Observations on the chaetotaxy of the legs in the free-living Gamasina (Acari: Mesostigmata). Bull Br Mus Nat Hist, Zool 10:277-303 Evans GO (1969) Observations on the ontogenetic development of the chaetotaxy of the tarsi of legs II-IV in the Mesostigmata (Acari). In Evans GO (ed) Proceedings of the II International Congress of Acarology, Budapest, 1967 :195-200 Evans GO, Till WM (1962) The Genus Dermanyssus De Geer (Acari : Mesostigmata). Bull Am Mus Nat Hist 13(5):273-293 Evans GO, Till WM (1964) A New Species of Dermanyssus and a redescription of Steatonyssus superans Zemskaya (Acari : Mesostigmata). Acarologia, 6(4):624-631 Evans GO, Till WM (1965) Studies on the British Dermanyssidae (Acari: Mesostigmata). Part I. External morphology. Bull Br Mus Nat Hist, Zool 13:249-294 Evans GO, Till WM (1966) Studies on the British Dermanyssidae (Acari: Mesostigmata). Part II. Classification. Bull Br Mus Nat Hist, Zool 14:109-370 Ewing HE (1923). The dermanyssid mites of North America. Proc US Natl Mus 62: 1-26 Ewing HE (1936) A short synopsis of the North America species of the mite genus Dermanyssus. Proc Entomol Soc Wash 38:47-54 Fain A (1993) A new species of the genus Dermanyssus DE GEER, 1778 (Acari: Dermanyssidae) from the nest of a bird Apus affinis in Rwanda. Bull ann Soc r entomolog Belg 129:163-168 Gu YM, Ting QY (1992) New species and new records of Dermanyssus from China (Acari: Dermanyssidae) Dongwu fenlei xuebao (Acta zootaxonomica Sinica)17(1): 32-36 Hirst S (1913). On three new species of Gamasid mites found on rats. Bull entomol res 4: 119-124 Krantz GW (1959) New synonymy in the Dermanyssidae Kolenati, 1859, with a description of a new species of Dermanyssus (Acarina, Dermanyssidae). Proc Entomol Soc Wash 61:174-178 Lindquist EE, Evans GO (1965) Taxonomic concepts in the Ascidae, with a modified setal nomenclature for the idiosoma of the Gamasina (Acarina: Mesostigmata). Mem Entomol Soc Can 47:1-64 Melville RV, Smith JDD (1987) Official lists and indexes of names and works in zoology. International Trust for Zoological Nomenclature, London, p. 308
59 Publication II
Moss WW (1966) Dermanyssus gallinoides n. sp. (Mesostigmata: Laelapoidea: Dermanyssidae), an Acarine Parasite of Woodpeckers in Western North America. The Canadian Entomologist 98:635-638 Moss WW (1967) Some new analytic and graphic approaches to numerical taxonomy, with an example from the Dermanyssidae (Acari). Syst zool 16:177-207 Moss WW (1968) An Illustrated Key to the Species of the Acarine Genus Dermanyssus (Mesostigmata: Laelapoidea: Dermanyssidae). J Med Ent 1:67-84 Moss WW (1978) The mite genus Dermanyssus : a survey, with description of Dermanyssus trochilinis, n. sp., and a revised key to the species (Acari: Mesostigmata: Dermanyssidae). J Med Ent 14(6):627-640 Moss WW, Radovsky FJ (1967) Dermanyssus hirsutus, a new name for Dermanyssus scutatus Ewing (Acari: Mesostigmata: Dermanyssidae). J Kans Entomol Soc 40:277 Moss WW, Mitchell CJ and Johnston DE (1970) New North American host and distribution records for the mite genus Dermanyssus (Acari: Mesostigmata: Dermanyssidae). J Med Ent 7:589-593 Nelson BC, Furman DP (1967) A new species of Dermanyssus from marine birds, with observations on its biology (Acarina: Dermanyssidae). Acarologia 9:330-337 Peterson A.P. http://www.zoonomen.net/avtax/frame.html Version 7.07 (2007.02.04) Radovsky FJ (1966) Revision of the Macronyssid and Laelapid mites of bats : outline of classification with descriptions of new genera and new type species. J Med Ent 3(1):93-99 Roy L, Chauve CM. (2007) Historical review of the genus Dermanyssus Dugès, 1834 (Acari: Mesostigmata: Dermanyssidae). Parasite. 14(2):87-100 Sheals GJ (1962) The status of the genera Dermanyssus, Allodermanyssus et Liponyssoides. In: Proceedings of the XIth International Congress of Entomology, Vienna, 1960 2:473-476 Strandtmann RW, Wharton GW (1958) A manual of mesostigmatid mites parasitic on vertebrates. Institute of Acarology, Contribution No. 4, University of Maryland, College Park, Maryland. Uchikawa K, Kitaoka S (1981) Dermanyssus nipponensis sp. nov. taken from Japanese green woodpecker indigenous to Japan (Acari: Mesostigmata) Natl Inst Anim Health Q 21(2):80-2 Uchikawa K, Takahashi M (1985) Contribution to mites of the genus Dermanyssus De Geer distributed in Japan (Acarina, Mesostigmata) Eisei Dobutsu (Jap J Sanit Zool) 36:3:233- 237 Valiente Moro, .C, Chauve C and Zenner L (2005) Vectorial role of some dermanyssoid mites (Acari, Mesostigmata, Dermanyssoidea) Parasite 12(2):99-109 Valiente Moro C, Chauve C, Zenner L.(2007) Experimental infection of Salmonella Enteritidis by the poultry red mite, Dermanyssus gallinae.Vet Parasitol. 146(3-4):329-36 Zeman P (1979). Dermanyssus carpathicus sp. n. (Acarina : Dermanyssidae), a new bird parasite from Czechoslovakia. Folia Parasitol 26:173-178 Zemskaya, A.A. 1971. [Mites of the family Dermanyssidae Kolenati, 1859, of the U.S.S.R. fauna]. Med. parazitol. parazit. bolezni 40: 709-717
60 Publication II
Figure 1 Dorsal shield of 10 from 20 randomly selected adult females of a cultured in lab population of D. gallinae. Setal terminology follows Lindquist and Evans (1965).
61 Publication II
Figure 2 Relative length of peritrema according to position of coxae. Each bar from left to right corresponds to length of peritrema in following species: D. faralloni D. trochilinis D. prognephilus D. grochovskae and D. hirundinis D. gallinae and D. gallinoides D. hirsutus D. triscutatus D. americanus D. alaudae and D. brevis D. transvaalensis and D. chelidonis
62 Publication II
Figure 3 Peritrema (scanning electron microscope) in D. gallinae
63 Publication II
is s e s s s s lu a n s i s u i s e s s u s s s k l s u l i i u n h i u s e m i c n s r ly s t a i i u e i e n n a p n v d s u n e e e i e i s a a i n r i h v n a a r c e o t o u l t n i e d m a d p u i s v o o t a i r i d r n d i e t r n i d u h s n l l a o i o fa i n s li i l u l h a n l u g n g v c c n l i c p p v t l u s i e l a s t r a d d a r n o e is o a r r o p e u ir ir a l i a me r h r r r a i n r a i w g a h lo p qu a b t g t g h a t n br w r B B ...... pr . .c . . . . .fa . . . .c . . . . D D D D D D D D D D D D D D D D D D D D D D D Fringillidae P Passeridae a Sittidae s Alaudidae s e Muscicapidae r Paridae i Hirundinidae f Corvidae o r Certhiidae m Laniidae e Motacillidae s Sylviidae Vireonidae Sturniformes Sturnidae Piciformes Picidae Strigiformes Strigidae Apodidae Apodiformes Trochilidae Accipitridae Ciconiiformes Alcidae Hydrobatidae Coraciiformes Meropidae ColumbiformesColumbidae Anseriformes Anatidae Galliformes Phasianidae Table 1 Host diversity for Dermanyssus species (from literature data). Species are in chronological order from left to right. Avian taxonomic groups were checked on http://www.zoonomen.net/avtax/frame.html.
64 4.3 Délimitation des espèces par une approche complémentaire (« total evidence approach »): publication III
a - Présentation L’extrême instabilité des caractères morphologiques traditionnels et le chevauchement interspécifique de leurs états, détaillés dans la publication II, laissent prévoir des mises en synonymie nombreuses à l’issue de la révision envisagée. Mais il faut vérifier l’absence d’isolement reproducteur entre les entités potentiellement synonymes. Les reconstructions phylogénétiques incluant plusieurs isolats par entité testée et portant sur des loci indépendants (mitochondriaux et nucléaires) offrent la possibilité de vérifier l’isolement reproducteur des entités spécifiques. La description de caractères morphologiques non décrits jusqu’à présent, ainsi que le recodage de certains des traditionnels permettraient en outre une exploration du genre Dermanyssus confrontant caractères moléculaires à caractères morphologiques.
4.3.a.1 Objectifs La troisième publication visait à clarifier les délimitations interspécifique dans le genre Dermanyssus, avec un focus sur le groupe gallinae en mettant à profit des techniques moléculaires, ainsi que de nouveaux caractères morphologiques (nouvellement décrits ou codés différemment).
4.3.a.2 Matériel et méthodes Suivant les recommandations de Samadi et Barberousse (2006), une approche a priori/a posteriori a été adoptée pour atteindre une délimitation stable des espèces. C’est-à-dire que l’alpha taxinomie, classification au niveau spécifique telle qu’elle est admise au moment de l’étude, doit être testée par des moyens indépendants, de manière à pouvoir se tourner vers elle a posteriori et confirmer/infirmer ses différentes composantes. Les moyens choisis pour le test de l’alpha taxinomie dans le genre Dermanyssus étaient une analyse phylogénétique comparative (matrice combinée vs matrice gène par gène, méthode Bayésienne vs maximum de parcimonie), impliquant à la fois caractères morphologiques et moléculaires, et à la fois gènes mitochondriaux et gènes nucléaires. L’intégration nouvelle de données moléculaires permettait d’espérer un tant soit peu de clarification dans ce genre à la morphologie apparemment peu informative. La confrontation de gènes nucléaires et mitochondriaux visait à détecter les éventuels flux de gènes entre entités supposées spécifiques (et ainsi remettre en question leur statut).
Quant aux caractères morphologiques, beaucoup des traditionnels paraissant impossibles à coder, nous avons recherché des caractères « nouveaux », utilisé un caractère peu utilisé, décrit par Moss (1968, K4, forme de la soie- al1) et recodé certains des anciens (ex : K9, longueur relative du péritrème). Une matrice de 46 caractères a ainsi pu être construite. Le travail de notation morphologique a été réalisé, outre sur les types (ADN indisponible), sur des individus dont les séquences d’ADN des régions ciblées ont pu être obtenues, de manière à ménager la possibilité du retour a posteriori sur la morphologie des individus testés, et, ainsi, sur l’alpha taxinomie.
65 4.3.a.3 Principaux résultats La réunion des types de 20 espèces parmi les 23 décrites au début de l’étude permit de décrire quelques rares caractères nouveaux (K6, K7, K26, K27, K31, K39, K40, K43), qui se présentaient pour la plupart comme discriminants entre groupe gallinae et groupe hirsutus (sous-genre Dermanyssus), et entre groupe gallinae et Microdermanyssus. Pratiquement aucun des caractères « nouveaux » ne semblaient permettre une distinction au sein du groupe gallinae. L’analyse phylogénétique de cette presque totalité des espèces du genre aboutit à un clade résolu pour les espèces hors groupe gallinae, tandis qu’une simple polytomie représente les relations des espèces du groupe gallinae entre elles et avec les autres.
L’analyse morphologique et moléculaire d’un échantillon de six de ces espèces, prélevées sur le terrain (ADN disponible), permit de mettre en évidence de flagrantes limites interspécifiques. Plusieurs populations ont pu être intégrées pour chacune des espèces testées (sauf deux, les deux seules hors groupe gallinae). Contre toute attente, des espèces a priori indiscernables entre elles laissèrent apparaître une nette isolation reproductrice. Le retour a posteriori sur l’ensemble des cuticules préparées pour l’observation microscopique et dont des séquences d’ADN ont pu être intégrées aux analyses permit de découvrir, parmi la multitude de caractères chaetotactiques variables intraspécifiquement, quelques portions de segments de pattes porteuses de soies à l’implantation régulière. Ainsi, D. hirundinis et D. longipes, dont les caractéristiques morphologiques semblaient trop incertaines, sont bel et bien des entités distinctes. Et D. apodis, une espèce nouvelle, fut révélée. La caractérisation moléculaire dans ces trois cas est confirmée par de subtils – mais à peu près stables – caractères morphologiques. Enfin, deux lignées basales de D. gallinae (L1 et L2) peuvent sembler isolées aussi. Mais s’il s’agit d’espèces à part entière, alors elles sont cryptiques, aucun élément morphologique discriminant n’ayant pu être mis en évidence.
b - Remarques sur la publication III
4.3.b.1 D. longipes : deux lignées différentes ? Les séquences d'ITS des deux populations testées de D. longipes (PAS et ENVLO83) présentent trois différences nucléotidiques l'une par rapport à l'autre, alors que les autres espèces ne présentent pour ainsi dire pas de mutations intraspécifiques (seulement 1 différence entre certains isolats de D. gallinae sur 1 site dans l’ITS1). L’isolat PAS provient d'un nid de moineau friquet (Passer montanus), et ENVL083 d'un nid de mésange charbonnière (Parus major). Or, ultérieurement, le séquençage d'une portion de gène codant pour la Tropomyosine (incluant un intron) a permis de relever des différences très importantes entre ces deux populations (publication V, p. 131 sqq.). En outre, un nouvel isolat provenant d'un nid de moineau friquet et un autre d'un nid de mésange charbonnière présentent exactement les mêmes différences respectives. Cela suggère deux lignées ou espèces différentes, inféodées chacune à un hôte différent, mais les prélèvements proviennent respectivement du même site, ce qui rend l'interprétation hasardeuse pour l'instant (différence liée à la localisation géographique ?). Toutefois, nous avons signalé dans la publication III comme caractéristique de D. longipes une séquence d'ITS attribuée par erreur à D. gallinae par Brännstörm et al. (2008). Or il s'agit de la séquence exacte de nos populations collectées sur mésange charbonnière. La séquence de Brännstörm et al. provient de plusieurs populations échantillonnées dans des nids de rousserolles (Acrocephalus sp) et de gobe-mouches (Ficedula sp) collectés en Suède. Nous n'avons malheureusement pas pu obtenir d'ADN de ces acariens afin d'en isoler la
66 Tropomyosine. Mais cela suggère que nous avons affaire là à la même lignée ou espèce que nos populations PAS et JBO108 (publication V) et que la localisation géographique est indépendante de la différence de trois nucléotides notée sur la séquence d'ITS. Enfin, les caractères morphologiques subtils notés sur les quelques individus de D. longipes intégrés à la publication IV se sont avérés, au vu des nouveaux échantillons, fortement instables entre isolats.
4.3.b.2 Un marqueur moléculaire abandonné : EF1-D Par ailleurs, au cours de la recherche et de la mise au point des marqueurs moléculaires pour cette partie de l'étude, nous avons isolé 25 séquences d'une région du gène codant pour l'elongation factor 1-D (EF1-D). Ces séquences correspondent aux isolats suivants: 14 isolats de D. gallinae, 2 de D. carpathicus, 3 de D. apodis, 2 de D. hirundinis, 1 de D. longipes (PAS) et 3 isolats d’outgroups. Nous avons finalement renoncé à l'utiliser dans le cadre de la reconstruction phylogénétique du genre Dermanyssus du fait d'une anomalie probablement témoin de paralogie : des individus appartenant à Ornithonyssus sylviarum (Macronyssidae) présentaient la même séquence, exactement, que D. gallinae. L'essai a été effectué à plusieurs reprises, à plusieurs mois d'intervalles, sur des extraits d'ADN d'individus d’O. sylviarum de deux provenances différentes (Bouches-du-Rhône, Rhône) et réalisés indépendamment, afin d'éliminer le motif de contamination. Ce marqueur nucléaire, s’il ne présentait pas une variabilité très élevée dans le genre Dermanyssus, était porteur tout de même d’un certain nombre de mutations ponctuelles entre espèces : 2 % de divergence entre D. gallinae et D. apodis, 3-4 % de divergence entre D. gallinae et D. carpathicus, 3 % entre D. apodis et D. carpathicus, 4 % de divergence entre D. apodis et D. hirundinis…Mais l’un des outgroups – distant de surcroît (O. sylviarum) – apparaît espèce sœur de D. gallinae, en position distale au sein du groupe gallinae dans les topologies obtenues. Avec D. longipes PAS, O. sylviarum présente exactement la même séquence que la plupart des D. gallinae (1 % de divergence entre isolats testés de D. gallinae). Trop contradictoire avec l’ensemble des autres données moléculaires et inconciliable avec les éléments morphologiques séparant clairement Macronyssidae de Dermanyssidae, cette information a été considérée comme inadéquate et mise de côté. Des cas de paralogie dans ce gène chez des arthropodes ont été notés à plusieurs reprises déjà (Djernaes & Damgaard, 2006). S’agit-il d’un cas de double copie avec disparition de l’une des deux chez certains taxa ? Pour l’heure, 25 séquences d’EF1-D obtenues avec les amorces DgEF1-Fn DgEF1-R, AcEF1-F et AcEF1-R sont disponibles dans la banque de gène internationale (numéros d’accès EMBL et séquences des amorces disponibles en Annexe 3).
4.3.b.3 Liponyssoides : genre introuvable ? Enfin, le genre Liponyssoides, seul autre genre avec Dermanyssus dans la famille des Dermanyssidae, n’a pas pu être intégré à la reconstruction phylogénétique de l’ensemble du genre, faute d’individu. Ce genre comprend actuellement 11 espèces, inféodées en général à des rongeurs et/ou des chiroptères, parfois à des oiseaux. Aucun des types demandés auprès d’institutions supposées les posséder n’a pu être obtenu. Trois des institutions susceptibles de posséder des types (holotypes ou paratypes) de Liponyssoides, le British Museum (Londres), Agriculture & Agrifood Canada (Ottawa), National Museum of Natural History (Washington) ont été contactées en vain : aucune trace des types, pas même de leur entrée en collection n’a pu être retrouvée, ni même de matériel non-type. Un examen des collections d’acariens du MNHN (Muséum National d’Histoire Naturelle, Paris ; M. Judson) n’a permis de trouver aucun individu de ce genre (type ou non type). Aucun des acarologistes contactés n’a pu fournir d’individu (liste des courriels d’acarologistes du British Muséum (NHM), J. Deunff, spécialisé dans la phylogénie des Rhinonyssidae, mésostigmates
67 parasites de chiroptères). En outre, nos efforts pour en isoler à partir de deux séries de prélèvements de terrain ont tous été vains : > 100 tubes contenant des acariens prélevés sur rongeur en Afrique, mis à disposition par Violaine Nicolas (MNHN) ont été examinés en vain (tous les spécimens étaient des Laelapidae (Mesostigmata) ; des photographies réalisées par Josyane Lips à partir d’échantillons de guano de chauves-souris de l’île de Vanuatu ont été visionnées en vain (absence de mésostigmate hématophage). Ashley P.G. Dowling a reçu une seule fois, en août 2007, en provenance d’Iran, un unique acarien prélevé sur rongeur, qui semblait correspondre à la description de Liponyssoides muris. Malheureusement, l’individu a été perdu au cours d’un déménagement. Moss (1978) ayant annoncé une révision du genre Liponyssoides en collaboration avec Strandtmann et Camin, il était possible que les types soient présents dans ses collections. Or Hans Klompen, (Museum of Biological Diversity, Columbus, Etats-Unis), qui, suite à l’acquisition par son institution de l’intégralité de la collection de Moss, y a eu accès, a pu retrouver certains types de Dermanyssus, qu’il m’a transmis, mais n’a trouvé aucune trace de Liponyssoides. Peut-être certains de ces types sont-ils dans des collections des collègues de Moss, Strandtmann ou Camin, décédés aujourd’hui. En bref, les espèces du genre Liponyssoides sont probablement peu fréquentes, ou très localisées et le mystère des types disparus reste entier.
68 Publication III Molecular Phylogenetics and Evolution 50 (2009) 446–470
Contents lists available at ScienceDirect
Molecular Phylogenetics and Evolution
journal homepage: www.elsevier.com/locate/ympev
Delimiting species boundaries within Dermanyssus Dugès, 1834 (Acari:Dermanyssidae) using a total evidence approach
L. Roy a,*, A.P.G. Dowling b, C.M. Chauve a, T. Buronfosse a a Université de Lyon, Ecole Nationale Vétérinaire de Lyon, Laboratoire de Parasitologie et maladies parasitaires, 69280 Marcy l’Etoile, France b University of Arkansas, Department of Entomology, Fayetteville, Arkansas, USA article info abstract
Article history: The genus Dermanyssus is currently composed of 24 hematophagous mite species and includes the Poultry Received 25 February 2008 Red Mite, Dermanyssus gallinae, a serious pest in poultry houses. Morphologically, Dermanyssus species fall Revised 12 September 2008 into two groups corresponding to Moss’ gallinae-group and to hirsutus-group + Microdermanyssus. Species Accepted 14 November 2008 of the gallinae-group exhibit high levels of morphological variability, and are nearly impossible to distin- Available online 27 November 2008 guish. Species of the second group display consistent characters and host associations and are easily distin- guishable. Species of the gallinae-group tend to be the major problems in poultry houses and it is unknown Keywords: whether D. gallinae is the only pest, or if there are numerous cryptic species present in the system. Dermanyssus Twenty species of Dermanyssus were tested phylogenetically based on 46 morphological characters. A Acari Mesostigmata subset of species, mainly of the gallinae-group, represented each by several populations, was sequenced Species boundaries for two mitochondrial and one nuclear gene regions. This allowed testing their specific status and their Phylogeny interrelationships based and on morphological and molecular characters. The molecular data was analysed Evolution separately and in combination with morphological characters. As expected, morphology did a poor job Taxonomy resolving relationships. Molecular data proved more informative. The resulting phylogenetic hypotheses brought some informa- tion about interrelationships among species of the gallinae-group showing a split into two main clades. The invasion of human managed environments seems to occur only in taxa within one of the two clades. The host spectrum seems to get enlarged in more derived taxa in the same clade. A delineation of six species within the gallinae-group is provided. Additionally, a key for morphological identification of these species is provided. D. gallinae appears to be the only pest in poultry houses, but is composed of several different and more or less strongly isolated lineages. A new species found from the black swift is described. 2008 Elsevier Inc. All rights reserved.
1. Introduction scribed since that date: Dermanyssus carpathicus Zeman, 1979, Dermanyssus nipponensis Uchikawa and Kitaoka, 1981, D. brevirivu- Genus Dermanyssus Dugès 1834 is currently composed of 24 lus Gu and Ting, 1992, D. wutaiensis Gu and Ting, 1992, Dermanys- hematophagous mite species, primarily parasitic on birds (Roy sus rwandae Fain, 1993 and Dermanyssus diphyes Knee, 2008. and Chauve, 2007; Knee, 2008).1 The Poultry Red Mite, Dermanyssus Dermanyssus is clearly defined compared to other genera due to gallinae (De Geer, 1778), is a serious pest in poultry houses, and other its roughly crescent-shaped and particularly short sternal shield Dermanyssus species have been shown to affect wild birds, such as and characteristic chelicerae. Chelicerae possess strongly reduced Dermanyssus prognephilus Ewing, 1933 on Purple martin chicks chelae and a strongly elongate, flattened and medially concave sec- (Moss and Camin, 1970) and Dermanyssus hirundinis (Hermann, ond segment, which forms something like a gutter and allows the 1804) on the offspring of House Wrens (Johnson and Albrecht, two chelicerae, once joined together, to form a tube through which 1993; Pacejka et al., 1996, 1998). Additionally, Clayton and Tomp- blood is sucked up (Phillis, 2006). On the other hand, species limits kins (1995) showed that D. gallinae can induce adult Rock Doves are not clearly defined and morphological characters traditionally Columba livia Gmelin, 1789 to spend less time incubating their eggs. used for diagnosis are highly variable within a single population No complete taxonomic review of the genus has been com- (Roy and Chauve, 2006) and even the same individual (bilateral pleted since 1978 (Moss, 18 species), and six species have been de- asymmetries). Consequently, some species other than D. gallinae may infest farms, which, if confirmed, may have serious conse- quences on control strategies. * Corresponding author. Fax: 33 478872577. In order to better define species limits, elucidate Dermanyssus E-mail address: [email protected] (L. Roy). phylogeny, and develop molecular tools for applied use, we have 1 Dermanyssus diphyes Knee, 2008 was published during paper revision, and therefore has not been included in present study. conducted a phylogenetic study of a part of the genus. The dataset
1055-7903/$ - see front matter 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.ympev.2008.11.012 Publication III L. Roy et al. / Molecular Phylogenetics and Evolution 50 (2009) 446–470 447 includes morphological characters, several of which have never ical characters. Sclerotized areas, usually bearing most of the phy- been examined for any study on Dermanyssus species relationships, logenetically informative characters, are strongly reduced in these and molecular data from ITS1 and 2 (plus some few bases of flank- species and are often asymmetric on a single individual. The dorsal ing regions of 18S and 28S rRNA,2 and including 5.8S rRNA), 16S shield also displays irregularities, including contours that are rRNA, and coding gene for Cytochrome oxidase subunit I (COI). asymmetric in an individual in almost all species of the gallinae- Dermanyssus species morphologically fall into two groups; those group and asymmetric setal patterns including numbers and posi- possessing a soft body adapted for sporadic and large engorgement tion. Additionally, leg chaetotaxy is highly variable intraspecifically with reduced shielding and slender legs (14 species) and those pos- (Evans and Till, 1962; Moss, 1978), a characteristic common among sessing a compact, more heavily sclerotized body with shorter, stou- mites that have formed parasitic associations (Evans, 1963). ter legs (9 species). Species possessing the soft-body type are the Such phenotypic variability not only makes species identifica- most common and most of them are nearly indistinguishable from tion difficult within a genus (Evans and Till, 1962; Moss, 1978), it each other; they constitute the gallinae-group: Dermanyssus antilla- also produces major problems for accurately coding morphological rum Dusbábek and Cˇerny´ , 1971, D. carpathicus Zeman, 1979, Derma- characters in a phylogenetic framework. Overall, this variability nyssus chelidonis Oudemans, 1939, Dermanyssus faralloni Nelson and has led to confusion regarding species limits and evolution within Furman, 1967, D. gallinae, Dermanyssus gallinoides Moss, 1966, D. hir- Dermanyssus and until this study, molecular characters have not undinis, Dermanyssus longipes (Berlese and Trouessart, 1889)(nomen been consulted. dubium), D. nipponensis Uchikawa and Kitaoka, 1981, D. prognephilus, The aim of the present study is to explore relationships between Dermanyssus transvaalensis Evans and Till, 1962, Dermanyssus tri- Dermanyssus species using a phylogenetic framework based on scutatus Krantz, 1959, Dermanyssus trochilinis Moss, 1978, D. wutai- morphological characters and between some species of the galli- ensis Gu & Ting, 1992. Several of these species have very large host nae-group using and morphological and molecular characters. ranges, in particular D. hirundinis and D. gallinae, which have been From obtained results, we plan (1) determining whether the galli- collected from numerous bird species, distributed across eight to nae-group includes distinct species or simply variants of popula- nine orders (Roy and Chauve, 2007). tions, (2) estimating whether the gallinae-group body type is Dermanyssus species possessing the second body type corre- primitive or derived and examine its adaptive significance and spond to Moss’ subgenus Microdermanyssus (Dermanyssus alaudae (3) evaluating host specificity of field collected species. (Schrank, 1781), Dermanyssus americanus Ewing 1922, D. brevirivu- lus Gu & Ting, 1992, Dermanyssus brevis Ewing, 1936, Dermanyssus grochovskae Zemskaya, 1961, Dermanyssus hirsutus, Dermanyssus 2. Material and methods passerinus Berlese and Trouessart, 1889, Dermanyssus quintus Vitzt- hum, 1921 and D. rwandae Fain, 1993) and members of his hirsu- In the aim of processing in a standardized manner, only adult tus-group (D. hirsutus Moss and Radovsky 1967, D. grochovskae and female mites have been used in this study. Adult females have D. quintus). All except D. quintus display a strong asymmetry in se- been selected as this is the only stage/sex described for all species. tae length between those situated centrally on the dorsal shield Adult males are less often found. Moreover, discriminant morpho- and those located on the perimeter. Several species possess con- logical characters appear to be mainly found in females. spicuous and distinctive morphological characters, such as a paired sclerotized porelike-structures on dorsum in D. alaudae, D. americ- 2.1. Methodology for delineation of species boundaries anus and D. brevis (Moss’ subgenus Microdermanyssus), a U-shaped row of large and deeply rooted setae on the opisthogaster in D. Primary hypotheses of alpha-taxonomy have been tested fol- quintus, and ventral neotrichy in the form of a cluster of elongate, lowing Samadi and Barberousse (2006) recommendations for help- simple setae laterad of the anal shield in D. hirsutus. These species ing in species delimitation. Our objective was to identify are more clearly distinguishable if compared to one another than reproductively isolated groups of organisms that warrant classifi- species of the gallinae-group on the basis of morphology. cation as distinct species by using phylogenetic tools. For such a Moreover, available data on these species suggest they are more purpose, successive validations of morphological characterization host specific than the gallinae-group, typically parasitizing a single with correlation to molecular information have been processed in bird family (Picidae for D. quintus and D. hirsutus, Alaudidae for D. order to test primary hypotheses provided by a-taxonomy. alaudae). However, D. grochovskae occurs on two bird orders, Pici- For testing primary hypotheses, two main actions have been formes and Passeriformes, and some of these species have been carried out. First, a comprehensive analysis of Dermanyssus phylog- found only once (D. brevis, D. brevirivulus), so the extent of their eny at the species level based on reference material has been car- host specificity is unknown. ried out, allowing us to obtain a set of discrete characters usable for Morphological differences between gallinae- and hirsutus- phylogenetic exploration. Second, partial exploration of Dermanys- groups have been suggested by Moss to be correlated to life-style. sus phylogeny involving various populations of several field col- Most Dermanyssus species are known to be nidicolous, climbing lected species, based on previously coded morphological onto the host only to obtain a meal before returning to their hid- characters and on molecular data. Several successive steps includ- ing-place in the host nest or roost. However, some species fre- ing comparisons between individual morphology and correspond- quently remain on the host for extended periods of time and can ing sequences was followed by phylogenetic analyses. Finally, deposit their eggs on its feathers. These species possess a morphol- some of the traditional species specific characters have been com- ogy more adapted to clinging onto the host rather than to running pared to the obtained phylogenetic hypotheses in order to assess around on it (e.g. D. grochovskae and D. quintus)(Moss, 1978). their actual utility (a posteriori feedback). Dermanyssus gallinae (gallinae-group) is of economic and veter- inary importance and it possesses highly polymorphic morpholog- 2.2. Morphological study
2.2.1. Taxon sampling 2 Abbreviations used: 16S, rRNA 16S; bp, base pairs; BPP, Bayesian posterior In the present study, Dermanyssus is represented by 20 of the probabilities; COI, cytochrome oxidase subunit I; ITS, rRNA 18S (partial sequence), currently recognized 24 species and one unidentified taxon internal transcibed spacer 1 (ITS1), 5.8S, internal transcribed spacer 2 (ITS2), 28S (partial sequence); Kx, morphological character n x; MP, maximum parsimony; OTU, (Appendix A). Type specimens of D. passerinus, D. brevirivulus and operational taxonomic unit; RSE, reference specific entities. D. wutaiensis were unavailable for examination (specimen dam- Publication III 448 L. Roy et al. / Molecular Phylogenetics and Evolution 50 (2009) 446–470 aged for D. passerinus and institution housing types for D. breviriv- Most samples come directly from wild bird nests, which have ulus and D. wutaiensis was unresponsive), and we did not find any been treated following the method of De Lillo (2001) with slight other reference specimens for these species. The type specimen of modifications. Overall, 327 nests were analyzed from 37 different D. longipes was also damaged but we were able to collect speci- bird species distributed across eight different orders. Bird taxon- mens in the field from the type locality near Avignon (France) as omy follows Peterson (2007). Due to their diversity and ubiquity, well as examine specimens in a Slovak collection (Fend’a, P., Come- passeriforms were the most represented host group, accounting nius University). D. diphyes has been described during revision of for 248 nests, with most nests distributed across four families present paper. Three species have been included as outgroups: (202 nests): Hirundinidae (46 nests), Parus sp. (Paridae; 120 nests), Ornithonyssus bacoti (Hirst, 1913) (Mesostigmata: Dermanyssoi- Alaudidae (14 nests), Passer sp. (Passeridae; 22 nests). The remain- dea: Macronyssidae), Haemogamasus hirsutus Berlese, 1889 (Meso- ing 46 passeriform nests were from 12 different species (less than stigmata: Dermanyssoidea: Haemogamasidae), Androlaelaps casalis ten nests/group). One species of apodiform is strongly present here (Berlese, 1887) (Mesostigmata: Dermanyssoidea: Laelapidae) and (Apus apus L., 1758, Apodidae; 52 nests). Other nest samples exam- Typhlodromus pyri Scheuten, 1857 (Mesostigmata: Ascoidea: Phy- ined were from the following bird groups: Columbiformes (13 toseiidae). The family Dermanyssidae also includes Liponyssoides, nests), Ciconiiformes (six nests), Strigiformes (three nests), Anser- but unfortunately no specimens were available for morphological iformes (two nests), Gruiformes (one nest), Piciformes (two nests). or molecular study, despite efforts of the authors (several collec- Additional collections of living mites have been made directly tions in which some types of Liponyssoides sp should have been from birds captured during bird-banding and/or bird care activi- deposited have been contacted, without any success). This has ties. This provided specimens from an additional species represent- forced us to only include distant outgroups of Dermanyssus. ing a ninth order: Coracias garrulus L., 1758 (Coraciiformes: Coraciidae). Three complementary populations have been obtained from other sampling activities (D. hirsutusADhirs, D. quintusADqui, 2.2.2. Character sampling D. hirundinisADhirun). Forty six morphological characters are included in the matrix (Appendix B and C). Due to very high variation (at the population 2.3.2. Taxon sampling level) of traditional chaetotactic characters (Roy and Chauve, Due to requirements of preserved specimens for DNA studies, 2006), stringent coding of such characters appeared impossible in our molecular dataset includes only those Dermanyssus species col- many cases. Therefore, we strongly reduced our reliance on such lected freshly into ethanol (or simply dried) by the authors or col- characters (only five traditional characters in present study), and laborators. Due to these constraints, very few Microdermanyssus + completely omitted leg chaetotaxy. We selected and coded 31 hirsutus-group have been included in this part of study: no Microd- additional morphological characters and ten morphometric ermanyssus and only two species of the Moss’ hirsutus-group (D. characters. quintus and D. hirsutus) have been collected. On the opposite, sig- nificant sampling of the gallinae-group has been included in the Five characters focus on chaetotaxy (K13, 14, 21, 34 and 35) of molecular dataset: four known and one unknown species of galli- anal and dorsal shields including the soft integument. nae-group are included. Because of the noted lack of discriminating Twenty one characters describe diverse parts of the body with characters found within the gallinae-group, several specimens five morphological characters describing soft integument (K10, were sampled from separate populations resulting in the inclusion 11, 18, 20 and 41), five describing shields (anal and dorsal of 45 gallinae-group OTUs in the combined molecular matrix, 29 of shields; K8, 23, 24, 25 and 36), two describing peritrema (K9 which are included in the total evidence analyses. and 42), one describing the palps (K1), five describing internal Only three of the four outgroups used in the morphological organs (K5, 6, 7, 19 and 45), two describing chelicerae (K32 analysis (O. bacoti, A. casalis, and T. pyri) were available fresh for and 33) and one describing cornicules (K46). molecular examination. Although efforts were made to collect Ten characters focus on the shape of some setae located on dor- specimens of Liponyssoides, all attempts were unsuccessful. sal shield, hypostome, legs, palps and soft integument of opis- thosoma (K4, 12, 15, 16, 17, 29, 30, 31, 43, and 44). 2.3.3. Four different a priori morphs on field collected species Ten characters use relative morphometry, six of which describe The key problems lie in the separation of the species of gallinae- dorsal, sternal, epigynial and anal shields (K2, 3, 22, 28, 37 and group due to variable characters within species and a general lack 38) and four of which describe the legs (K26, 27, 39 and 40). of discriminating characteristics across species. Therefore, a first and rough examination of material led to delimitation of four a pri- ori morphogroups. Only species of gallinae-group are dealt with 2.2.3. Phylogenetic analysis based on morphological data here, as members of the ‘‘Microdermanyssus + hirsutus-group” are For the phylogenetic analysis, all characters were treated as easily defined. Initial examination focused on numerous popula- unordered and unweighted. A heuristic analysis was performed tions across Europe and resulted in separation into the following under the parsimony criterion using PAUP* 4.0b10 (Swofford, four morphogroups based upon preliminary successive compari- 2001) with TBR branch swapping and 10,000 random additions sons of sequences and morphological data: DG-morph (type popu- saving all most parsimonious trees. Heuristic searches in TNT lation SK (Table 1)—al1 of palp genu lanceolate, sternal shield with (Goloboff et al., 2008) were used to obtain relative Bremer (Golob- a deep central concave neckline, dorsal shield narrower than podo- off and Farris, 2001) and bootstrap support values. TNT searches soma), GO-morph (type population GO1—al1 of palp genu lanceo- recovered the same topology and tree length as PAUP, but calcula- late, sternal shield without a sharp postero-medial neckline, dorsal tion of support values is much more efficient in TNT. shield as wide as podosoma, anterior pair of setae in hypostomal parallelogram larger than in D. gallinae), RQ-morph (type popula- 2.3. Molecular exploration tion RQ—al1 of palp genu spine-like) and DL-morph (type popula- tion PAS—similar to DG-morph, but with a slightly different anal 2.3.1. Biological material shield, proportionally longer and slightly more angular—cf. above). Dermanyssus specimens were collected from November 2005 to The following populations were sequenced during this study May 2008, mainly in France (some in USA), using two sampling however, not all were included in the final combined analyses, typ- methods due to the different lifestyles found within the genus. ically because of missing data (Table 1). Publication III L. Roy et al. / Molecular Phylogenetics and Evolution 50 (2009) 446–470 449
Table 1 Taxonomic sampling and EMBL accession numbers for each sequence. Each record corresponds to one mite population belonging to a single species sampled from a single nest or a single bird. Locality and host information are also provided. Species a priori morph Pop Accession No. Country, French Host Context code department 18S–28S rRNA 16S rRNA mt-COI D. carpathicus RQ-morph RQ AM903316 AM921903 AM921876 France, 42 Phoenicurus phoenicurus Nest, near a human (Passeriform) house, alt ca. 1500 m RQ-morph 5. AM903314 AM921901 AM921873 France, 42 Parus major (Passeriform) Nest box RQ-morph Veol AM921871 France, 69 Parus major (Passeriform) Nest box RQ-morph Parm AM903315 France, 69 Parus major (Passeriform) Nest box RQ-morph JBO59 AM930882 AM921902 AM921870 France, 84 Parus major (Passeriform) Nest box in an apple orchard RQ-morph LC10A FM179367 Parus major (Passeriform) Nest RQ-morph LR20A FM179368 Parus sp. (Passeriform) Nest RQ-morph JMC10 AM943018 AM943021 France, 62 Parus major (Passeriform) Nest D. gallinae DG-morph SK AM903303 AM921887 AM921856 Denmark Gallus gallus (Galliform) Layer farm *Special lineage DG-morph SB AM921858 France, 69 Gallus gallus (Galliform) Little amateur one (L1). hen house DG-morph Chab AM931074 AM921886 AM921857 France, 01 Gallus gallus (Galliform) Layer farm DG-morph Fa1 AM931072 AM921884 AM921853 Norway Gallus gallus (Galliform) Layer farm DG-morph Fa2 AM931071 AM921883 AM921852 Norway Gallus gallus (Galliform) Layer farm DG-morph PO1 AM903302 AM921854 Poland Gallus gallus (Galliform) Layer farm DG-morph PO2 AM921914 AM921855 Poland Gallus gallus (Galliform) layer farm DG-morph DR AM931073 AM921885 Spain Fringillidae (Passeriform) Cage DG-morph ROL1 AM903304 AM921910 AM921864 France, 13 Coracias garrulus (Coraciiform) On adult birds DG-morph ROL2 AM903305 AM921911 AM921865 France, 13 Coracias garrulus (Coraciiform) On young birds at nest DG-morph Woodp AM903301 AM921890 AM921863 France, 69 Dendrocopos major (Piciform) On wild adult female bird DG-morph CANIT AM903308 AM921909 AM921877 Italy Serinus canarius (Fringillidae: Breeding facility Passeriform) DG-morph LB07 AM921866 France, 18 Delichon urbica (Passeriform) Nest DG-morph LB18 AM930889 AM921908 AM921867 France, 18 Delichon urbica (Passeriform) Nest DG-morph JBO51 AM930885 AM921879 France, 84 Parus major (Passeriform) Nest box in an apple orchard DG-morph Percnobis AM943020 France, 07 Neophron percnopterus Nest (Ciconiiform) DG-morph LC* AM903306 AM921891 AM921859 France, 26 Columba livia (Columbiform) Breeding facility, rural country DG-morph COL* AM903307 AM921892 AM921875 France, 69 Columba livia (Columbiform) On adult bird DG-morph PI* FM179378 FM179375 AM921860 France, 13 Columba livia (Columbiform) City center, coming from nest inside a flat DG-morph GO8* AM921893 France, 30 Apus apus (Apodiform) Nest DG-morph JGC1* AM921861 France, 26 Tyto alba (Strigiform) Nest D. hirsutus hirsutus-group ADhirs AM931077 AM921912 AM921878 USA, MI Colaptes cafer (Piciform) On bird D. hirundinis DG-morph HR AM903300 AM921888 AM921872 France, 69 Hirundo rustica (Passeriform) Nest DG-morph OC AM903312 AM921889 AM921862 France, 38 Delichon urbica (Passeriform) On young birds at nest DG-morph ADhirun AM931076 AM921913 AM921881 USA, MI Tachycineta bicolor (Passeriform) On bird DG-morph HIR1 FM179379 FM179366 France, 85 Hirundo rustica (Passeriform) Nest DG-morph CHOV AM943019 FM179369 France, 72 Hirundo rustica (Passeriform) Nest in a barn D. longipes DG-morph PAS AM903310 AM921904 AM921869 France, 84 Passer montanus (Passeriform) Nest DG-morph ENVL08-3 FM179377 FM179374 FM179365 France, 69 Parus sp. (Passeriform) Nest box D. quintus hirsutus-group ADqui AM931075 AM921882 USA, MI Picoides villosus (Piciform) On bird D. apodis GO-morph MAR AM945880 AM921899 AM921880 France, 69 Apus apus (Apodiform) On young bird fallen from nest GO-morph GO1 AM903299 AM921894 France, 30 Apus apus (Apodiform) Nest GO-morph GO10 AM921895 France, 30 Apus apus (Apodiform) Nest GO-morph GO12 AM903309 France, 30 Apus apus (Apodiform) Nest GO-morph GO15 AM903313 AM921896 France, 30 Apus apus (Apodiform) On young birds at nest GO-morph GO16 AM903313 France, 30 Apus apus (Apodiform) On young birds at nest GO-morph GO26 AM921900 France, 30 Apus apus (Apodiform) Nest GO-morph GO36 FM179371 France, 30 Apus apus (Apodiform) Nest GO-morph GO44 AM921898 France, 30 Apus apus (Apodiform) Nest GO-morph GO46 AM921897 France, 30 Apus apus (Apodiform) Nest GO-morph GO54 AM930888 AM921874 France, 30 Apus apus (Apodiform) Nest GO-morph GO58a FM179370 France, 30 Apus apus (Apodiform) Nest O. bacoti outgroup Ob AM903318 AM921905 FM179677 ? Rodents From a lab strain in MNHN (O. Bain, Lab of Parasitology) A.casalis outgroup ACA AM903317 AM921907 AM921868 France, 69 — Breeding facility T. pyri outgroup TPYR FM179376 FM179373 FM179364 ? — From a lab strain in Supagro, Montpellier (S. Kreiter) Publication III 450 L. Roy et al. / Molecular Phylogenetics and Evolution 50 (2009) 446–470
gallinae-group: GO-morph, 11 populations; DG-morph, 25 pop- undergone DNA extraction were compared to these vouchers to ulations; RQ-morph, 7 populations; DL-morph, 2 populations. determine the usefulness of the DNA voucher cuticles. The protein- ase K digestion did not appear to have any adverse effects on the 2.3.3.1. D. longipes. Dermanyssus longipes was deemed nomen dubi- cuticule and all characteristics necessary for morphological exam- um due to damaged type material and absence of additional re- ination remained intact. DNA was extracted following procedures cords since description. One population has been isolated with in the Qiagen QIAamp DNA Mini Kit. When possible (most cases), the DL-morph and assigned to D. longipes, due to two morpholog- two to three separate specimens from each tested population were ical characteristics and to geographic locality. The anal plate is extracted and sequenced. slightly more elongated and subrectangular and the dorsal shield has a subapical shrinkage and ends with a rounded apex. Of course 2.3.5.2. DNA amplification and sequencing. PCRs were separately it was considered that this might simply be another polymorphic performed in order to amplify two mitochondrial gene regions character, but many specimens from the same nest (Passer mont- (part of COI gene and of 16S) and one nuclear region (fragment anus, South of France) and other from P. domesticus, in Slovakia, 18S-ITS1-5.8S-ITS2) in either a Biometra TGradient or a MWG AG provided exactly the same characteristics. Repetition of these anal Biotech Primus 96plus thermal cycler in typical buffer containing plate characters not only within a single population, but also in a 2 ll of template DNA, 2.5 U of Taq polymerase, 10 nmol of dNTPs, geographically distant second population led us to conclude this 20 pmol of each primer and a variable volume of 50 mM MgCl2 is a reliable morphological difference from other specimens with depending of the target gene in accordance to Table 2 in a final vol- the DG-morph. Moreover, this population has been collected near ume of 50 ll. After an initial denaturation step (95 C) for 10 min, the same locality (20 km) and from the same host genus as D. long- followed by 40 cycles of: 20 s at 95 C (denaturation), 30 s at the ipes type material (Berlese and Trouessart, 1889). annealing temperature specified for each primer set (Table 2), and 90 s at 72 C (extension). A final extension step was carried 2.3.4. Character sampling out for 10 min at 72 C. Several primers have been designed for Three gene regions have been selected for sequencing; two amplification in various species and are provided in Table 2. mitochondrial markers (COI and 16S) and one nuclear marker Negative and positive controls were run with each round of (ITS). The molecular dataset includes a total of 1524 bp for each amplification. PCR products were checked by electrophoresis in a taxon (cf. accession numbers in Table 1) and sequence data was 1% agarose gel. Depending on the brightness of the band either aligned using MAFFT (Katoh et al., 2005) with the L-INS-i iterative additional PCRs were run on the original template or reamplifica- refinement option on the MAFFT server at http://align.bmr.kyushu- tions of the original PCR product were performed. PCR reamplifica- u.ac.jp/mafft/online/server/. MAFFT with the L-INS-i option has tions using same primers were assessed on 1 ll of product and PCR shown to be the most accurate and consistent method for se- conditions were as follows: initial denaturation at 95 C for 3 min quences (Wilm et al., 2006; Carroll et al., 2007). followed by 20 cycles of: 20 s at 95 C, 45 s at specified annealing temperature (Table 2), and 2 min at 72 C. A final extension step 2.3.5. Amplification and sequencing of DNA was carried out for 45 min at 60 C. In both cases, a total of four 2.3.5.1. Morphological preparation and DNA extraction. DNA was ex- reactions were run for each taxon sample and the resulting ampli- tracted from individual mites by cutting the cuticle at two points fications were pooled in order to obtain enough DNA for sequenc- on the opisthosoma and pushing most of the internal elements ing. The four PCR tubes from each sample were pooled together, out. This was done with a sterile pipet tip in the appropriate com- submitted for electrophoresis in a 1% agarose gel, and PCR products mercial buffer containing proteinase K (Qiagen) and digestion was were excised from the gel and purified using the Macherey-Nagel performed at 70 C for 19–30 h. The cuticle was separated from the Nucleospin Extract-II kit. Purified PCR products were sequenced DNA mixture and mounted as a voucher and for microscopic obser- by Genoscreen (France, Lille) using a 96-capillary sequencer vation. Specimens slide mounted directly from alcohol having not ABI3730XL.
Table 2 Primer sequences and key parameters for PCR conditions.
Gene Primer name Primer sequence Primer annealing T ( C) MgCl2 (mM) Comments ITS-forward DgITS-F 50-AGAGGAAGTAAAAGTCGTAACAAGG-30 48 3 RhITS-F* 50-AGAGGAAGTAAAAGTCGTAACAAG-30 ITS-reverse DgITS-R 50-CCTTAGTAATATGCTTAAATTCAGG-30 Amplicon including only 44 bp 28S RhITS-R* 50-ATATGCTTAAATTCAGGGGG-30 Amplicon including only 37 bp 28S AmITS-R2 50-GTTAGACTCCTTGGTCCGTGTTTCA-30 Amplicon including more than 600 bp 28S mt-COI-forward COF1 50-ATCGGAGGATTCGGAAACTG-30 52.5 3.3 COF1bis 50-CTGCACCTGACATGGCTTTCCCAC-30 CO1F4 50-CACCTGACATGGCTTTCCCACGAT-30 CO1LCF 50-GAAAGAGGAGCAGGCACTGG-30 Works well with special lineage one CO1RQF1 50-GAAAGAGGAACAGGAACAGG-30 Specific to D. carpathicus, D. longipes and D. hirundinis mt-COI-reverse RQ-COI-R 50-CCAGTAATACCTCCAATTGTAAAT-30 COIDpro-R 50-GTTGGGATTGCAATAAT-30 Works well with special lineage one COIGOR 50-GTTGGAATTGCAATAAT-30 Works well with D. apodis ObCOIF-rev 50-GTGGGAATHGCAATAAT-30 16S-forward Rh16S** 50-GCTCAATGATTTTTTAAATTGCTG-30 54 3.3 Dg16SF 50-TGGGTGCTAAGAGAATGGATG-30 16S-reverse Rh16S** 50-CCGGTCTGAACTCAGATCATG-30 Dg16SR 50-CCGGTCTGAACTCAGATCAAG-30
* Primer sequence from De Rojas et al. (2002). ** Primer sequence from De Rojas et al. (2001). Publication III L. Roy et al. / Molecular Phylogenetics and Evolution 50 (2009) 446–470 451
2.3.6. Phylogenetic analyses based on combined morphological and Two other Bayesian runs were completed in order to test the molecular data effects of 3rd position change in the COI analysis as well as Phylogenetic analyses with Maximum Parsimony (MP) were missing data in the molecular only analysis. In many analyses run for the total evidence analysis, a combined molecular only involving COI, the majority of change or divergence between analysis, and individual analyses of morphology and the separate taxa or populations is found in the highly variable 3rd position. gene regions. Heuristic searches were carried out in PAUP* Results based solely on 3rd position change are often criticized 4.0b10 (Swofford, 2001) with TBR branch swapping and 10,000 and considered based upon random data. This analysis excluded random additions saving all most parsimonious trees. Heuristic 3rd positions from the COI single gene analysis to determine if searches in TNT (Goloboff et al., 2008) were used to obtain relative any presently supported groups would still exist in the final tree. Bremer (Goloboff and Farris, 2001) and bootstrap support values. The analysis was run with all the same parameters as previous For Bayesian analyses the total evidence data set (morphology Bayesian runs and an appropriate model determined by and molecules), the combined molecular only data set, and the MrModeltest. individual genes were all run using MrBayes (Huelsenbeck and The primary reason for using a subset of taxa in the combined Ronquist, 2001; Ronquist and Huelsenbeck, 2003). MrBayes differs analyses was to eliminate a large amount of missing data that from other programs in allowing partitions within the data set to would have to be incorporated in order to include every population implement different models of evolution, presumably allowing sequenced for at least one gene region. In order to test whether for a more realistic analysis of the data. Models of evolution were these missing data would have had an impact on the final analyses, applied to individual molecular partitions and determined for each all taxa sequenced for at least one region were included in a final gene by MrModeltest (Nylander, 2004) for Bayesian analyses. In ‘‘all-taxa” molecular only dataset. The Bayesian analysis was run the total evidence Bayesian analysis, the following models were under the same parameters as the other molecular only combined applied to each partition: (1) Morphology used the standard dis- analysis. crete model (appropriate for likelihood approximations of morpho- logical datasets; Lewis, 2001) and assumed gamma-shaped rate 3. Results variation; (2) ITS used GTR + i (proportion of invariable sites with- out a gamma distribution); (3) COI used GTR + C + i; and 4) 16S 3.1. Comprehensive phylogenetic reconstruction based on morphology used GTR + C. Each of the models for the molecular partitions alone was determined in MrModeltest using Akaike information criterion (Akaike, 1974). Parameters within each model were not specified The MP heuristic analysis of 25 taxa and 46 morphological char- (or fixed) and MrBayes was left to estimate these independently acters resulted in 12 most parsimonious trees (L = 129, CI = 0.4264, for each partition from the data during analysis. All analyses in RI = 0.6085) and the strict consensus is represented in Fig. 1. The MrBayes included two independent runs, each consisting of four monophyly of tested species of Dermanyssus is supported by a very chains and 5,000,000 generations. Appropriate burnins were short sternal shield (K3), a strongly atrophied third cheliceral seg- determined based on stationarity being reached through the use ment (K32), a strongly elongate and foliate second cheliceral seg- of Tracer v1.4 (Rambaut and Drummond, 2007). ment (K33) and membranous cornicules (K46).
Typhlodromus pyri Androlaelaps casalis 52 6 1.00 Haemogamasus hirsutus 57 1.00 Ornithonyssusb acoti 33 D. trochilinis D. longipes gallinae-group (part) D. hirundinis 93 3 32 46 D. carpathicus 1.00 6 3 3 17 40 D. grochovskae 65 0.44 D. hirsutus hirsutus-group 6 39 72 D. quintus 1.00 D. rwandae
39 40 D. brevis 53 Microdermanyssus 0.44 60 D. alaudae 83 0.13 D. americanus 0.44 D. nipponensis D. transvaalensis D. prognephilus D. chelidonis D. faralloni gallinae-group (part) D. gallinae D. gallinoides
50 D. antillarum 0.42 D. triscutatus D. apodis
Fig. 1. Strict consensus tree of 12 most parsimonious trees (L = 129, CI = 0.4264, RI = 0.6085) using matrix of 46 morphological characters. The numbers below nodes refer to the relative Bremer support and numbers above refer to bootstrap percentages from 1000 replicates. Additionally, mapping of the main morphological synapomorphies is figured by white (character state 0), grey (character state 1) and black (character state 2) dots, labeled with corresponding character number. Publication III 452 L. Roy et al. / Molecular Phylogenetics and Evolution 50 (2009) 446–470
Within Dermanyssus, D. trochilinis is a sister to all other Derma- The clade G contains the RQ-morph, DL-morph, some of the popu- nyssus species, although support at this node is moderate (0.44 rel- lations with DG-morph, and unlike the MP analysis, the popula- ative Bremer, 83% bootstrap). The remaining species share a tions with GO-morph and the hirsutus-group are basally resolved strongly unresolved basal node and the clade grouping Microder- and showing affinities to this first clade. Each of the morphs found manyssus and the hirsutus-group is the only group with any kind in this large clade represents a monophyletic group of populations. of relative Bremer support and includes eight species (D. alaudae, Within this clade, the populations with GO-morphs are most basal D. americanus, D. brevis, D. grochovskae, D. hirsutus, D. quintus, D. and strongly supported (1.0 Bayesian Posterior Probabilities (BPP)), rwandae). Monophyly of the group is supported within Dermanys- followed by the hirsutus-group (0.74 BPP), the RQ-morphs (1.0 sus by one apomorphy: the proportions of tibia I (K39, with an evo- BPP), then the populations with DL-morph (1.0 BPP), and finally lution in distal OTUs in D. alaudae + D. americanus). Moreover, the a clade of DG-morphs (1.0 BPP). The other main clade (B) consists character state 2 in K6 (shape of the principal pore on the post- of the remaining populations with DG-morph and shows little se- stigmatic element) is an apomorphy of this group within Derma- quence variation across the clade. This clade, like in the MP analy- nyssus, but there is a homoplasy with one outgroup (H. hirsutus sis, is very poorly supported (0.55 BPP) indicating the basal shares same character state in K6). Note that the hirsutus-group relationships within this clade are questionable. All other relation- does not appear to be monophyletic here, the interrelationships ships within the clade B corroborate the results of the MP analysis. between the three species keep unresolved. It should be interesting The results of the ‘‘all-taxa” matrix represented in Appendix E to test the monophyly of this group using molecular data in a fur- show that the same groups are resolved, albeit with more mem- ther study. D. carpathicus, one of the gallinae-group species, ap- bers, and for all intents and purposes provides the same topology pears as sister group of above clade (0.44 relative Bremer; 65% as the previously run molecular-only combined analysis. The large bootstrap). Two synapomorphies support this new clade: character amount of missing data apparently had no effect on the final re- state 1 in K17 (relative length of setae on dorsal side of femur I) sults, and in fact because the same major groups emerged, would and 2 in K40 (proportions of genu I), with an evolution in distal indicate how strongly the individual genes converge on the same OTUs in Microdermanyssus, (transition to the third state 1). As for hypothesis. other species, interrelationships in the gallinae-group keep unre- solved (forming a large polytomy), except for the sister group rela- 3.2.2. Total evidence matrix tionship of D. antillarum and D. triscutatus (0.42 relative Bremer; The MP heuristic analysis of 34 taxa and 1570 characters (46 50% bootstrap). morphological K, 643 bp of ITS, 540 bp of COI, 341 bp of 16S) re- sulted in 288 most parsimonious trees (L = 1426 CI = 0.6452 3.2. Combined analyses on the subset of field collected species RI = 0.7634). Strict consensus (Fig. 3A) resulted in a hypothesis identical to the molecular only result. Molecular information has been obtained from one to three The Bayesian analysis once again resulted in a well resolved tree genes in 46 different populations (Table 1). Obtained topologies (Fig. 3B) very similar to the molecular only tree (Fig. 2B). One major are shown in Figs. 2A and 3A. Distribution of pairwise divergence difference is the placement of the hirsutus-group. In the molecular among tested populations in the three tested genes is provided only tree, the hirsutus-group was sister to the RQ-morph + DL- in Fig. 4. morph + DG-morph clade (clade H in Fig. 2B), whereas in the total evidence results, the hirsutus-group is sister to the RQ-morph clade 3.2.1. Combined molecular matrix and overall (clade I’ in Fig. 3B), support values are much higher. All The MP heuristic analysis of 34 taxa and 1524 characters other groupings remain the same between the two results. (643 bp of ITS, 540 bp of COI, 341 bp of 16S) resulted in 576 most Removing 3rd positions from the COI analysis produced a less parsimonious trees (L = 1357 CI = 0.6478 RI = 0.7641). A strict con- resolved phylogenetic hypothesis (Fig. 5) than the full COI analysis, sensus (Fig. 2A) resulted in two main clades within Dermanyssus. as expected. However, the following major groups were still recov- The first groups most of the DG-morph populations (clade B) with ered: GO-morph (1.0 BPP), RQ-morph (1.0 BPP), RQ + DL-morphs fairly strong support (0.77 relative Bremer; 98% bootstrap). Within group (0.76 BPP), GO + DG + hirsutus + DL + RQ-morphs group this clade, several internal clades are strongly supported. Of them, (0.82 BPP), and two DG-morph subgroups from the main DG- two successive sister clades (F and E) group strongly together three morph clade typically recovered (COL + JGC + LC + PI, corresponding and two populations respectively (1.0 relative Bremer and 100% to clade F: 0.88 BPP and JBO51 + LB07 + LB18, corresponding to clade bootstrap scores) and distinctively from the other DG-morphs E: 0.95 BPP). These results indicate that while major change is and each other. The second clade groups populations with RQ- present in the 3rd position, informative change (information) still morph, DL-morph and the remaining DG-morph (clade I) although exists in the more slowly evolving 1st and 2nd positions. with very little support at the basal nodes. The individual morphs are monophyletic group with good support, including RQ-morph (J, 3.3. Single gene analyses 0.55 relative Bremer; 100% bootstrap), DL-morph (N, 0.90 relative Bremer; 100% bootstrap), and the remaining DG-morphs (L, 0.45 COI: The MP heuristic analysis of 41 taxa and 540 characters of relative Bremer; 98% bootstrap). The remaining groups are basally COI resulted in 36 most parsimonious trees (L = 699 CI = 0.5622 unresolved and show no affinities to either of the previously men- RI = 0.8216). The strict consensus resulted in a similar topology tioned clades. This includes the hirsutus-group (D. hirsutus and D. to the results of the combined analyses, although more unresolved. quintus) which is entirely unresolved and the populations with Most notably is the lack of resolution between the DL-morph and GO-morph, which form a monophyletic group with strong support the DG-morphs, normally resolved in the combined analyses. (1.0 relative Bremer; 100% bootstrap) but with relationships to The Bayesian results for COI are more resolved than the MP re- other Dermanyssus unresolved. sults and the only difference between the COI results and the com- In the Bayesian analysis different models of evolution were bined results is the separation of D. hirsutus and D. quintus. implemented for each partition (gene region) in the molecular 16S: The MP heuristic analysis of 34 taxa and 341 bp of 16S re- dataset and resulted in a more resolved tree than the MP run, sulted in 28 most parsimonious trees (L = 309 CI = 0.6796 although many characteristics of the two analyses are similar. RI = 0.8467). The strict consensus resulted in a similar result to The topology obtained from MrBayes (Fig. 2B) is almost completely the combined analyses, although slightly less resolved. Unlike the resolved and results also in two major clades of Dermanyssus. COI analysis, relationships of RQ-, DL- and part of DG-morphs are Publication III L. Roy et al. / Molecular Phylogenetics and Evolution 50 (2009) 446–470 453
Typhlodromus pyri A Ornithonyssus bacoti Androlaelaps casalis 93 CANIT 1.00 ROL1 Chab 86 D DR 1.00 Fa1 SK 59 51 74 ROL2 C 0.50 0.50 Woodp 1.00 D. gallinae 95 Fa2 98 1.00 PO2 B 100 0.77 E JBO517 1.00 LB18 100 LC 100 A 1.00 F 67 COL 0.78 1.00 PI 100 GO1 1.00 GO54 D. apodis MAR D. hirsutus D. quintus ADhirun 98 L HIR1 0.45 68 M CHOV D. hirundinis 64 HR K 0.44 0.43 OC 100 58 N ENVL083 D. longipes I 0.90 PAS 0.22 100 JMC10 100 J 0.45 RQ D. carpathicus 0.55 100 JBO59 0.36 A5
B Typhlodromus pyri Ornithonyssus bacoti Androlaelaps casalis 0.99 5 1.00 JBO59 J D. carpathicus JMC10 1.00 RQ 0.80 ADhirun I 1.00 HIR1 L CHOV D. hirundinis M 0.96 K 1.00 HR 0.60 H 0.96 OC ENVL083 N D. longipes 0.93 1.00 PAS G D. hirsutus 0.74 D. quintus 1.00 GO1 GO54 D. apodis MAR CANIT 1.00 ROL1 A Chab D DR 0.89 CANIT Fa1 ROL1 0.97 SK Chab C Fa2 DR D. gallinae 1.00 PO2 0.54 Fa1 ROL2 SK Woodp 1.00 1.00 B Fa2 0.55 JBO517 E 0.81 PO2 LB18 ROL2 LC 0.55 0.99 Woodp F COL 1.00 JBO517 0.1 PI LB18 1.00 LC COL 0.53 PI 0.01
Fig. 2. Molecular combined analysis using 1570 bp from cytochrome oxidase subunit I, rRNA 16S and rRNA 18S–28S, including ITS1, 5.8S and ITS2. (A) Maximum parsimony criterion, PAUP 4.0. Strict consensus of 576 most parsimonious trees (L = 1357 CI = 0.6478 RI = 0.7641). The numbers below nodes refer to the relative Bremer support and numbers above refer to bootstrap percentages from 1000 replicates. (B) Bayesian analysis from 5,000,000 generations using partitioned data and independent models of evolution for each partition. Numbers on nodes refer to Bayesian posterior probabilities. resolved, but with RQ-morph grouped with part of DG-morph, and ITS: The MP heuristic analysis of 35 taxa and 643 bp from ITS re- with DL-morph as a sister group (different than results found in the sulted in 1000 most parsimonious trees (L = 413 CI = 0.8184 combined analyses). Finally, interrelationships within the second RI = 0.7706). The strict consensus resulted in a conspicuously less main clade are less resolved than in previous analyses. Bayesian re- resolved topology than in the combined analyses. The monophyly sults are basically identical to the MP results, showing the same of populations with RQ-morph appears strongly supported (1.0 rel- relationships. ative Bremer; 92% bootstrap) however no other relationships are Publication III 454 L. Roy et al. / Molecular Phylogenetics and Evolution 50 (2009) 446–470
Typhlodromus pyri A Ornithonyssus bacoti Androlaelaps casalis 95 CANIT 1.00 ROL1 Chab 91 D DR 1.00 Fa1 SK 60 64 66 ROL2 C 0.57 1.00 0.57 Woodp D. gallinae 96 Fa2 100 1.00 PO2 B E 1.00 97 JBO517 1.00 LB18 100 F LC 60 2 3 4 6 32 34 43 1.00 COL 100 A 1.00 PI 1.00 46 100 GO1 GO54 D. apodis 1.00 4 6 17 43 MAR D. hirsutus D. quintus 1 6 17 ADhirun 98 L HIR1 CHOV 26 40 0.28 D. hirundinis 77 K HR 0.28 OC 100 127 N 74 1 26 27 40 43 ENVL083 D. longipes I 1.00 PAS 0.16 98 4 17 JMC10 100 J 0.28 RQ D. carpathicus 1.00 100 JBO59 1.00 A5
B Typhlodromus pyri Ornithonyssus bacoti Androlaelaps casalis 0.99 5 1.00 J JBO59 D. carpathicus 0.89 JMC10 I 1.00 RQ D. hirsutus 0.98 D. quintus 0.99 H ADhirun 1.00 L HIR1 0.53 D. hirundinis M CHOV HR 0.92 K 1.00 G 1.00 OC .1.00 ENVL083 N D. longipes PAS 1.00 GO1 GO54 D. apodis MAR CANIT ROL1 A D Chab 0.90 DR CANIT Fa1 1.00ROL1 0.98 SK Chab C Fa2 DR D. gallinae 1.00 PO2 Fa1 ROL2 0.55 SK Woodp 1.00 100 Fa2 B 0.60 JBO517 0.83 E PO2 LB18 ROL2 LC 0.60 0.99 Woodp F 1.00 COL JBO517 PI LB18 1.00 LC 0.1 COL 0.55 0.01 PI
Fig. 3. Total evidence analyses using 46 morphological characters and 1570 bp from cytochrome oxidase subunit I, rRNA 16S and rRNA 18S–28S, including ITS1, 5.8S and ITS 2. (A) Maximum parsimony criterion, PAUP 4.0. Strict consensus of 288 most parsimonious trees (L = 1426 CI = 0.6452 RI = 0.7634). The numbers below nodes refer to the relative Bremer support and numbers above refer to bootstrap percentages from 1000 replicates. Additionally, mapping of the main morphological synapomorphies is figured by white (character state 0), grey (character state 1) and black (character state 2) dots, labeled with corresponding character number. (B) Bayesian analysis from 5,000,000 generations using partitioned data and independent models of evolution for each partition. Numbers on nodes refer to Bayesian posterior probabilities. Moreover, triangular signs indicate populations found in a human-shaped environment, internal color corresponding to different bird groups (grey triangle, pigeons breeding facilities, white triangle, fringillids breeding facilities, black triangle, layer hen houses). Publication III L. Roy et al. / Molecular Phylogenetics and Evolution 50 (2009) 446–470 455
A B 35 30 16S 30 COI 25 25 20 20 L3 L2 L1 15 15 L3 L2 L1 10 10
5 5
0 0 0 100 200 300 400 500 0 100 200 300 400 500 pair number pair number CD 30 ITS Pair members 16S COI ITS 25 intra-RSE L3 L3 0-1% 20 inter-RSE L2 L2 2-8% RSE + D. longipes L2 L2 2-8% 15 RSE + D gallinae non-LB L2 L2 2-8% L3L3 + L2 L2 L1 D. gallinae except special lineage 1 L3 L2 1% 10 D. gallinae non-LB + special lineage 1 L3 L2 2% D. gallinae special lineage 1 + LB L2 L2 2% 5 D. gallinae non-LB + D. apodis L2 L2 1% outgroup + ingroup L1 L1 L1 0 0 100 200 300 400 500 pair number
Fig. 4. Distribution of percentages of pairwise divergence among populations of the eight OTUs used in molecular analyses. RSE = reference specific entities (cf. discussion, § species), ie D. carpathicus, D. hirsutus, D. quintus. L1, L2, L3 = hierarchical levels 1, 2, 3 noted on trees are discussed in the text.
Typhlodromus pyri Androlaelaps casalis Ornithonyssus bacoti CANIT Chab Fa1 Fa2 PO1 PO2 D. gallinae (part) ROL1 ROL2 0.81 SB SK Woodp ENVL08 3 D. longipes 0.76 PAS LC10A 1.00 0.97 5 JBO59 0.66 D. carpathicus 1.00 LR20A 0.64 JMC10 0.52 RQ 0.94 Veol 0.54 D. hirsutus D. quintus ADhirun 0.82 HIR1 D. hirundinis 0.60 CHOV 0.73 HR OC 1.00 GO54 GO36 D. apodis 0.81 GO58a MAR COL 0.88 JGC 0.1 LC PI D. gallinae (part) 0.95 JBO51 7 LB07 4 LB18 0.05
Fig. 5. Bayesian analysis of the COI matrix, excluding 3rd positions of codons. Publication III 456 L. Roy et al. / Molecular Phylogenetics and Evolution 50 (2009) 446–470 discernable. Bayesian analysis provided a slightly more resolved CANIT, Chab, DR, Fa1, Fa2, JBO51-7, LB18, PO2, ROL1, ROL2, SK, result including a weakly supported clade of DG- and DL-morphs Woodp) group together with 0.60 BPP in Bayesian and 1.0 relative as in other results and a strongly supported clade of RQ-morph. Bremer and 100% bootstrap, in MP combined analyses and corre- Additionally, the DG-morph clade F forms a strongly supported spond to D. gallinae. monophyletic group. Populations with GO-morph (GO1, GO54, MAR) group together As individual units, none of the gene regions show high resolv- together with 1.0 BPP in Bayesian and 1.0 relative Bremer and 100% ing power or large amounts of useful phylogenetic data. However, bootstrap in MP combined analyses and correspond to D. apodis n. as a combined unit the resolving power greatly increases and sp. described below. understandable relationships emerge, thus reinforcing the utility Moreover, observations on the distribution of pairwise diver- of total evidence approaches. gences between populations in mitochondrial genes allowed detection of three main levels of hierarchy represented in the trees 3.4. Species boundaries (Fig. 4), which partially corroborate above cladistic delineations. Separation between level 1 (20-25% in COI, 22–31% in 16S and In all three single gene analyses, as in the total evidence and 19–25% in ITS) corresponds to separation between the ingroup combined analyses, the same populations typically group together, and outgroup and level 2 (8–18% in COI, 9–16% in 16S and 0–8% and it’s the more internal nodes where disagreements are found. in ITS) to separation between species (RSE, see below). Level 3 Mitochondrial gene analyses resolve rather similarly to each other, (0–6% in COI and 0–7% in 16S) is internal within ingroup (i.e. con- with less resolution intra-morph in 16S than in COI-based topolo- cerns differences within species). In the nuclear gene, level 3 does gies. The ITS-based topology is less resolved, but still recovers the not generate a conspicuous gap and part of levels 2 and 3 are over- RQ-morph group. lapping each other (1–2% pairwise divergence). On the whole, phylogenetic analyses of the total evidence and On the other hand, non-hirundinis populations with DG-morph molecular-only matrices, recovered a monophyletic grouping of do not clearly group together with strong support values, except the different expected species of the gallinae-group with strong for some subsets of populations. Populations of clade F appear support values (bootstrap values from 98 to 100% in all analyses; clearly grouped together in the three single gene analyses and sep- Table 3). One interesting result, however, was the consistent split- arated from other D. gallinae populations in the ITS analysis (sister ting of the DG-morphs into two clades. In the analyses, the GO- group and to D. gallinae and to the GO-morph group). Other D. gal- morph reveals an isolated entity along with the following species linae, together with this subset, form a monophyletic clade with level delineations: the GO-morph clade in ITS single gene analyses, but with very weak bootstrap support (Table 3). Moreover, the populations of – Populations with DL-morph (ENVL08, PAS) group together with clade E in Figs. 2 and 3 form a strongly supported clade in COI anal- 1.0 BPP in Bayesian and 1.0 relative Bremer and 100% bootstrap, ysis, but group in a weakly supported clade and together with the in MP combined analyses and correspond to D. longipes. population CANIT in ITS single gene analysis (Table 3). – Populations with RQ-morph (RQ, 5, JBO59, JMC10) group together with 1.0 BPP in Bayesian and 1.0 relative Bremer and 3.5. A posteriori observation of some traditional characters 100% bootstrap, in MP combined analyses and correspond to D. carpathicus. Observation of traditional descriptive characters from the re- tained cuticles of sequenced individuals confirmed the strong var- Population with DG-morph split into two different clades, cor- iability of most of these traits. Closer examination in reference to responding to at least two species. These four populations (ADhi, phylogeny does indicate that some subtle characters do exist that Hir1, HR, OC) group together with 1.0 BPP in Bayesian and 0.28 rel- may be useful for species distinction (cf. diagnostic characters in ative Bremer and 98% bootstrap, in MP combined analyses and cor- the key for identification below §4.5). Here are results on a poste- respond to D. hirundinis. These fifteen populations (COL, LC, PI, riori tested characters.
Table 3 Support values for monophyly of species with several populations and other groups in the three single genes analyses (Maximum parsimony and bayesian) Nm = Not monophyletic. a priori Group 16S COI ITS morphs BA MP BA MP BA MP Number of Bayesian Branch support Number of Bayesian Branch support Number of Bayesian Branch tested posterior for monophyly tested posterior for monophyly tested posterior support for populations probabilities (% bootstrap/rel. populations probabilities (% bootstrap/rel. populations probabilities monophyly Bremer’s) Bremer’s) (% bootstrap/rel. Bremer’s) DG-morph D. gallinae except 11 0.58 43/0.25 14 0.58 NM 12 NM NM lineage 1(L1) DG-morph D. hirundinis 3 1.00 100/1.00 5 1.00 98/1.00 5 NM NM RQ-morph D. carpathicus 3 0.97 96/1.00 7 1.00 100/0.78 4 1.00 92/1.00 DG-morph D. gallinae special 4 0.99 100/0.93 4 1.00 100/1.00 3 1.00 69/1.00 lineage 1(L1) GO-morph D. apodis 7 1.00 99/1.00 4 1.00 100/1.00 3 NM NM DG-morph (D. gallinaeLB; 1 — — 3 1.00 100/1.00 1 — — JB051)= clade E DG-morph (D. hirundinis; D. longipes) 5 NM NM 7 0.92 NM 7 0.99 NM DG-morph D. gallinae 15 0.81 89/0.88 18 0.73 86/0.90 15 NM NM DG+GO- (D. gallinae; D. apodis) 22 NM NM 22 NM NM 18 0.87 39/0.29 morph Genus Dermanyssus 30 0.89 67/0.80 36 1.00 89/0.92 29 1.00 100/1.00 Publication III L. Roy et al. / Molecular Phylogenetics and Evolution 50 (2009) 446–470 457
3.5.1. Dorsal shield chaetotaxy states that this condition is present in some species and absent The dorsal shield has very rough and irregular contours in many in some others, indicating its utility in species identification. How- specimens, and j1 and s1 are sometimes situated on the shield and ever, upon investigation, we found several cases where individuals other times off the shield. Several specimens have been found with found in the same aggregation exhibited body types with and identical DNA sequences but variable position of such setae. without these shoulders or pronotal scutella. Additionally, when This is especially conspicuous in D. gallinae where 21% of se- tested genetically with the three gene regions used in this study, quenced individuals had j1 clearly off shield, 16% clearly on, 53% all produced identical sequences. Moss observed that the ‘‘promi- along the edge and 11% exhibited clear asymmetry (off on one side, nent shoulders” of D. gallinae were formed ‘‘by the fusion of the on on the other side). In the clade B, 67% of sequenced individuals dorsal shield with the platelet” (Moss, 1966). These platelets in fact had j1 along the edge of the shield and 33% with clear asymmetry. seem to be present in most species (K41), but can be fused or un- In other species, the number of sequenced individuals is lesser, so fused to the dorsal shield within a single population of one species. fewer variations have been noted. However, below is an overview Furthermore, many cases of bilateral asymmetry were discovered of noted variations: in this study, especially in D. carpathicus (25% of sequenced indi- viduals) and in D. gallinae (9.5%). In D. carpathicus, 57% of sequenced individuals had j1 off and 43% with clear asymmetry. 4. Discussion In populations GOn and MAR, which slightly differ morphologi- cally and strongly group together, 67% of sequenced individuals On the whole, the morphology-based analysis does not provide had j1 off shield and 33% with clear asymmetry. adequate information to determine species delineation nor phyloge- netic relationships within the gallinae-group. Additionally, this anal- In some species, such as, D. hirundinis and D. longipes, 100% of ysis only brings some information about possible relationships sequenced individuals had j1 off the shield without variation. within Microdermanyssus + hirsutus-group (cf. below). On the other Moreover, variations in the total number of setae present on the hand, analyses including molecular data, either with MP or Bayesian, dorsal shield have been noted in every species: 21–25 setae in D. provide valuable information regarding delineation of some species hirundinis, 17–24 setae in D. carpathicus, 21–30 setae in D. gallinae, and species’ relationships (cf. below). All converge on similar results, 20–25 setae in D. longipes, 25–29 setae in populations GOn and with more or less resolution. The least resolved topologies result MAR. from ITS single gene analyses (MP, Bayesian) indicating that most of the resolving power is found in the mitochondrial genes. Total evi- 3.5.2. Leg chaetotaxy dence and combined molecular only datasets offer much more reso- Leg chaetotaxy is highly variable in parasitic mite species lution in the gallinae-group than the morphological. (Evans, 1963), and especially within Dermanyssus (cf. Evans and The present study took into account mitochondrial and nuclear Till, 1962; Moss, 1968). Within Dermanyssus, we found intraspe- genes. As it has previously been observed at similar levels (Moore, cific and intra-individual variation across many species. In order 1995; Springer et al., 2001; Shaw, 2002), our results suggest that to determine if there are any evolutionary tendencies or patterns the tested nuclear gene (ITS) has less resolving power than the in leg chaetotaxy, we mapped these characters onto the molecular tested mitochondrial genes in recovering relationships within Der- phylogenetic hypothesis. Because of high mobility of setae on the manyssus. Therefore, our species delineation and evaluation of rela- legs, traditionally annotated setae ad, pd, av and pv have been con- tionships are mainly based on mitochondrial data. It is recognized densed down to ventral (v) and dorsal (d) notation. We also noted that this could be misleading for inferring species phylogenies due anterior lateral (al) and posterior lateral (pl) setae on each taxon. It to the haploid character of mitochondrial origin, however, single was not possible to compare sides with each other in all individu- gene analyses do not produce results highly contradictory to the als due to occasional problems in cuticular treatment during DNA total evidence and combined analyses. extraction or mounting (cf. Material and methods above). Dermanyssus gallinae proved to be the most variable of all spe- cies with individual asymmetry (number of setae differing from 4.1. Species one side to the other) observed in 50% of sequenced individuals (on femur I v, in genu I and II d, v and pl and in genu IV d and While the combined dataset does provide some interesting pl). In symmetrical individuals, intraspecific variation was found information about the gallinae-group phylogeny, the main purpose to be very high. Variation in the number of setae in homologous of incorporating molecules that tend to sort out closely related spe- area between individuals has been noted on femur I v in 18% of se- cies, or even distinct populations, was to determine if the gallinae- quenced individuals, in 11% on femur I pl, in 19% on genu II and III group actually constitute a number of morphologically similar spe- v, in 6% on genu IV d and v. This variation is found not only be- cies or if they represent one homogenous population of D. gallinae. tween populations, but also within single populations. Clearly leg This is important in terms of dealing with pest species of domestic chaetotaxy is not a phylogenetically informative character and ap- birds, which to this point have been continuously identified as D. pears amazingly plastic during the development of D. gallinae. gallinae. In other sequenced species of the gallinae-group, asymmetries (number of setae differing from one side to the other) were noted 4.1.1. Distribution of multi-population a priori morphs in 50% or more for sequenced individuals of D. hirundinis (mainly In order to reveal species boundaries within morphologically ventral faces of femur I, genu I, II and IV, and al and dorsal face similar entities, it is important to include several geographically in genu II). Additionally, variation between individuals of D. hir- distant populations in phylogenetic analyses (Monaghan et al., undinis was found in 14% of sequenced individuals on genu III al 2005). In the present study, several geographically distant popula- and in 50% on genu II v, genu III d, and genu IV d. tions (from various places in France) from the four a priori morphs have been tested in combined analyses (three genes and morphol- 3.5.3. Pronotal scutella ogy, three genes only) and in single gene analyses (Appendix D)of Moss (1966) identifies the pronotal scutella as the anterior por- both mitochondrial and nuclear sequences (GO-morph: 10 popula- tion of the dorsal shield not rounded, but with two somewhat tions; DG-morph: 45 populations; RQ-morph: 5 populations). Pop- acute and laterally pointed angles, suggestive of shoulders. He ulations of GO-morph and RQ-morph are, respectively, grouped Publication III 458 L. Roy et al. / Molecular Phylogenetics and Evolution 50 (2009) 446–470 together in strongly supported clades (Table 3 and Figs. 2 and 3), these populations are almost identical to that of other D. gallinae whereas DG-morph is clearly separated in two distinct clades: pop- populations, except for one site (common to CANIT). ulations HR, OC, CHOV, HIR1 and ADhirun in the clade L (Figs. 2 and On the other hand, populations of the clade F (Figs. 2 and 3) ap- 3), and the other DG-morphs (Table 1) in the clade B (Figs. 2 and 3). pear morphologically identical to D. gallinae populations, but re- solve monophyletic in all analyses. They are divergent from other 4.1.2. D. carpathicus, D. hirsutus and D. quintus, clearly characterized D. gallinae in all analyses, and may represent a cryptic species. Thus described species, as reference specific entities single gene analyses resulted in similar topologies, and grouped Three species appeared clearly characterized early in the study: these populations in strongly supported clades both in mitochon- D. carpathicus (sharp pairwise divergence with other entities, very drial and nuclear genes (Table 3), with identical ITS sequences be- little divergence between geographically distant populations with- tween each other and differing by 2% from D. gallinae populations. in the species, phylogenetically grouped together with strong sup- There is likely no gene flow between populations COL, LC, PI, GO8 port in all analyses (Table 3), slight morphological differences but and JGC1 (from various environments, with geographical distances clear divergence); D. hirsutus (pairwise divergence with other enti- between them from about 100 to 300 km) and other tested popu- ties, sharp morphological divergence); and D. quintus (pairwise lations of D. gallinae, but since there are no clear diagnostic mor- divergence with other entities, sharp morphological divergence). phological characters, these populations may at best represent Of course, respective monophylies of D. quintus and D. hirsutus recent speciation or cryptic species. As Heethoff et al. (2007) con- have not been tested, due to lack of additional populations, but cluded when studying potentially cryptic species of the oribatid their DNA sequences are very divergent from each other and other mite Platynothrus peltifer Koch, 1839, we have decided to make populations and morphological characterization is obvious. For all no decision regarding species status until more biological informa- these reasons, these three specific entities will be used in the pres- tion is obtained. Thus, as recommended by DeSalle et al. (2005),it ent study as references for species status and are referred to as ref- is necessary to get corroboration for more than one line of evidence erence specific entities (RSE) in order to have a comparison of the for delineation of a new taxon. Here, DNA is the only line of evi- distribution of pairwise divergence percentages (Fig. 4). Pairwise dence. In the case of a cryptic species, as no morphological clues divergences located in hierarchical level 2 in Fig. 4 correspond to are available (even subtle as in some species of Tectocepheus in interspecific divergence. Laumann et al., 2007), some geographical or ecological data is nec- essary to corroborate the DNA evidence. Future studies will try and 4.1.3. D. hirundinis and D. longipes obtain additional samples from pigeons of various geographical Populations of D. hirundinis also form distinct clades separate origins and various types of environment (breeding facilities, urban from other Dermanyssus displaying large genetic divergence and nests), as an ecological common trait seems to be the host group some diagnostic morphological characters have been noted in the (cf. below). In present paper, this entity will be refered to as D. gal- a posteriori feedback described in Section 4.3. D. longipes is a sister linae special lineage one. to D. hirundinis populations in all analyses, and this group as a whole is distinct from other Dermanyssus. There is a slight excep- 4.1.5. Does population Woodp belong to D. gallinoides? tion in 16S single gene Bayesian analysis, where it appears sister Moss (1966) described D. gallinoides as follows: (1) no promi- to a clade including D. carpathicus and D. hirundinis. These results nent shoulders (anterior part of the dorsal shield not rounded, provide confidence that D. hirundinis is a unique species. Moreover, but with two somewhat acute and laterally pointed angles, sugges- as tested populations of D. longipes not only group together in all tive of shoulders), (2) small platelet on the soft integument on each topologies and are separate from D. hirundinis in 16S gene analysis, side of the dorsal shield, (3) dorsal shield scaling smooth, (4) j1 and but also it is 11.2% (16S) and 9.4% (COI) different from D. hirundi- s1 off the dorsal shield (D. gallinae: j1 always and s1 generally on nisAD, it is apparent that it also represents a good species concept. the dorsal shield), (5) epigynial pores off the shield, (6) tibia IV pl Additionally, slight morphological differences between D. longipes with 2 setae (instead of one in D. gallinae), (7) genu II pl, III al and all other Dermanyssus species have been noted. The anal plate and IV al with two setae (opposed to only one seta in D. prognephi- is slightly more elongated (relative measures) and more or less lus), (8) peritreme extending only to the middle or anterior margin subrectangular and the dorsal shield has a subapical shrinkage of coxa II rather than to the middle of coxa I (different from D. pro- and ends with a quite rounded apex. In most cases, monophyly gnephilus). Upon examination of numerous populations of D. galli- of tested populations in phylogenetic analyses supported not only nae and now recognizing the great amounts of variability in many by this amount of genetic divergence, but also by a few morpholog- of these characters, the sixth argument appears to be the only valid ical characters typically constitute unique species status. It would one. In terms of host associations, D. gallinoides, has been found on be interesting to include more than two populations of this species several different species of Picidae, a group not normally associated in the future in order to firmly fix the specific status. with D. gallinae. We collected one specimen originally identified as D. gallinae (labelled Woodp population) from an adult female 4.1.4. Non-hirundinis populations with DG-morph Dendrocopos major (Picidae) that exhibits several characteristics The results from the analysis of the combined dataset indicate resembling D. gallinoides. that what had been identified by the authors as D. gallinae is not as The shape of dorsal shield fits arguments one and two depend- clear and does in fact group into several different lineages. These ing on the observed side (bilateral asymmetry) and position of j1 is populations do not group together in a strongly supported clade in asymmetric (one side on, the other side off shield, 4th argument). single gene analyses either (Table 3). Moreover, pairwise divergence Of course, depending on which side of the body you look at, these between them is in some cases in level 2 and in some others in level 1. character states either direct you to D. gallinoides or D. gallinae. Ti- Populations LB18, JBO51 (clade E in Figs. 2 and 3) and LB07-4 bia IV pl only has one seta, like in D. gallinae, but femur I pl only resolve monophyletic in single gene COI analysis (94% bootstrap), posesses one seta (instead of two in all other D. gallinae). Geneti- but not in ITS analysis, where LB18 and JBO51 branch with popu- cally, population Woodp falls within the other D. gallinae speci- lation D. gallinae CANIT (Table 3). This incongruence between mito- mens indicating even more variation within that species, and if chondrial and nuclear genes suggests that, even if these entities are this does in fact represent what has been called D. gallinoides,it partially isolated from each other, there is some gene flow between suggests that this should be synonymised with D. gallinae. How- them and D. gallinae populations. This implies that it has to be ever, many more specimens would be necessary to make this deemed belonging to D. gallinae. Additionally, ITS sequences in decision. Publication III L. Roy et al. / Molecular Phylogenetics and Evolution 50 (2009) 446–470 459
4.1.6. D. apodis n. sp long (47–65 lm), lateral pair 26 lm long (24–31 lm), posterior Strongly supported monophyly of populations GOn and MAR pair 26 lm (26–28 lm). Capitulum 96 lm (96–137 lm) long (from along with several subtle morphological characters separate these its basis to apex of palp coxae), 151 lm (143–166 lm) wide basally populations from other species of the gallinae-group. These popu- and 117 lm (104–130 lm) wide distally, (i.e. between lateral mar- lations constitute an entity which appears to be of specific status gins of palp coxae’ apex). and which we describe here under the name D. apodis. Setae al1 of palp genu lanceolate. Anterior hypostomal setae Female (holotype) (Fig. 6A–B): pair wider than other gnathosomal setae. Gnathosoma. Length of setae: anterior pair of hypostomal setae Idiosoma. 840 lm (735–1050 lm) long and 494 lm (420– 37 lm long (range with 5 paratypes 31–39 lm), central pair 57 lm 693 lm) wide. Dorsum: dorsal shield length 714 lm (646–
Fig. 6. D. apodis n. sp. (A) Venter of an adult female (holotype). (B) Dorsal shield of an adult female (holotype). (C) Venter of an adult male. (D) Dorsal shield of an adult male. Publication III 460 L. Roy et al. / Molecular Phylogenetics and Evolution 50 (2009) 446–470
798 lm), width 286 lm (262–291 lm) at midlevel. Its anterior Moreover, branch lengths are much more important between this margin with a concave slit between the two anterior pores. These entity and others than within the cluster (Monaghan et al., pores are often anteriorly located and separated from the shield. 2005), which highlights the unique status of this particular hierar- (Fig. 6B). Relative length of dorsal shield more than twice the podo- chy in the tree (idem in D. carpathicus and D. hirundinis). soma area bounded by coxae (367 lm, 4 par. 346–451 lm) in Among the 24 species of Dermanyssus described so far, only length. Relative width of dorsal shield almost as wide as the podo- three species have been recorded from some Apodiformes: D. hir- soma area bounded by coxae, with lateral margin running across undinis (on swifts, family Apodidae), D. rwandae (on swifts, family each coxa. Shape of ultrastructural network on dorsal shield Apodidae) and D. trochilinis (on hummingbirds, family Trochilidae). slightly differing on anterior part and on middle and posterior part: Closely related to D. gallinae (and especially the subset of pop- grooves delimiting rather short areas (about as wide as long) in ulations COL, JGC1, PI, LC), D. apodis morphologically differs from anterior part and longer areas around the middle of dorsal shield, it mainly by the concave slit between the two anterior pores of which seem to converge toward posterior part via longitudinal dorsal shield (continuously rounded convex margin in D. gallinae). axis. No major difference of length between central/peripheral se- Moreover, the pronotal scutella are not fused to dorsal shield (usu- tae of dorsal shield (series j4-6 and z5/j2, z2, z4 and s4). Pronotal ally fused in D. gallinae) and the anterior pair of hypostomal setae scutella present, separated from dorsal shield. Venter (Fig. 6A). is slightly wider than in D. gallinae. It also clearly differs from D. Sternal shield 18 lm (18–29 lm) long and 148 lm (122– hirundinis by the concave slit between the two anterior pores of 171 lm) wide. Genito-ventral shield 254 lm (234–260 lm) long dorsal shield (continuously rounded convex margin in D. hirundi- and 140 lm (119–140 lm) wide at midlevel. Oviporal flap nis) and by some elements of leg chaetotaxy (pl of genu II and III 130 lm (109–169 lm) long. Anal shield 153 lm (148–174 lm) with 2 setae in D. apodis,1inD. hirundinis). It also clearly differs long and 148 lm (137–156 lm) wide, with anterior margin’s out- from D. trochilinis by the concave slit between the two anterior line very irregular. Post-stigmatic trachea (which extends posteri- pores of dorsal shield (continuously rounded convex margin in D. orly from stigmata): principal pore large (ca.3 setae base), a trochilinis), by the relative width of setae of anterior hypostomal hole surrounded by a large raised chitinous ring, forming some- pair (about as wide as other hypostomal setae) and by the absence thing like a neck. of pronotal scutella. Legs. Tibia I 99 lm long (94–101 lm) and 60 lm wide (52– Etymology. The species name is derived from the specific name 65 lm). Tibia II 78 lm (73–75 lm) long and 52 lm wide (48–60 lm). of host and is the genitive form of the word. Genu I 99 lm long (94–99 lm) and 73 lm wide (62–75 lm). Genu Material examined. II 80 lm long (75–78 lm) and 68 lm wide (57–70 lm). Chaetot- axy of legs: Genu I 2–5/3–2; Genu II 2–4/2–2; Genu III 2–4/1–2; – Individuals ex nests and adult birds of A. apus (Apodiformes: Genu IV 2–4/1–1. Variations in paratypes (6 examined): Genu I: Apodidae), Nîmes, France (Gard), June–July 2007: one paratype 2–3/3–2 and two with a bilateral asymmetry on ven- tral face. Genu II: two paratypes with a bilateral asymmetry on Holotype female (one individual of population GO54, (n MNHN ventral face. Genu III: one paratype with a bilateral asymmetry on Ac1111a, cf. Table 1). Seven paratype adult females from following ventral face, two not determined (legs III lost). Genu IV: one paratype populations: GO1 (n MNHN Ac1112), GO15 (n MNHN Ac1113b), 1–4/1–0, another with a bilateral asymmetry on ventral face. GO46 (n MNHN Ac1116a and n MNHN Ac1111b), GO54 (n Nucleic acids. Several amplicons from the three tested genes MNHN Ac1111b), GO59 (n MNHN Ac1117a and n MNHN have been sequenced for different populations of D. apodis (acces- Ac1117b); 2 paratype deutonymphs: GO15 (n MNHN Ac1113a), sion numbers in Table 1: holotype belongs to population GO54 and GO16 (n MNHN Ac1113c); 2 paratype adult males: GO44 (n paratype females to populations GO1, GO15, GO16, GO46, GO59, MNHN Ac1115), GO59 (n MNHN Ac1117c). MAR). All obtained ITS and 16S sequences were exactly the same. Nest samples from which these mites have been isolated and Only two bases were different between sequences of COI obtained samples directly caught from birds in this locality have kindly been from the holotype population (holotype individuals and another provided by G. Gory (Muséum d’Histoire naturelle de Nîmes and individual found from the same swift nest) and the paratype pop- Centre de Recherche sur la Biologie et les Populations d’Oiseaux). ulation MAR (0.3% divergence). Male (paratype) (Fig. 6C and D) – Individuals ex a young individual of A. apus, Francheville, France Gnathosoma. Length of setae: anterior pair of hypostomal setae (Rhône), July 17th, 2007: 26 lm long (18 lm in a second male paratype), central pair 47 lm long (31 lm), lateral pair 33 lm long (not visible in second para- 3 paratype adult females from population MAR (n MNHN type), posterior pair 14 lm (13 lm). Capitulum 75 lm (75 lm) Ac1114a, n MNHN Ac1114b, n MNHN Ac1114c). long (from its basis to apex of palp coxae), 127 lm (98 lm) wide basally and 109 lm (91 lm) wide distally, (i.e. between lateral G. Lallemand sampled these mites during care activity in the margins of palp coxae’ apex). Spermatodactyl 96 lm long. Centre de Soins aux Oiseaux Sauvages du Lyonnais. Setae al1 of palp genu lanceolate. Anterior hypostomal setae The holotype and paratype series are deposited in the Museum pair not wider than other gnathosomal setae; note: thinner ante- National d’Histoire Naturelle, Paris, France. rior hypostomal setae in male than in female appears to be a sexual dimorphism present also in D. gallinae. 4.2. Different rates of evolution Idiosoma. 746 lm (693 lm) long and 525 lm (451 lm) wide. Dorsum: dorsal shield (Fig. 6D) length 651 lm (599 lm), width The amount of genetic differentiation within and between spe- 316 lm (260 lm) at midlevel. Podosoma area bounded by coxae cies varied depending on the gene, however, we did not find any 340 lm (315 lm) in length. Venter (Fig. 6C). Sternigenitoanal evidence of intrapopulation variation among any of the gene re- shield 578 lm (536 lm) long and 187 lm (176 lm) wide. The gen- gions sampled. The nuclear marker (ITS region) provided few but ital orifice is located on the anterior margin of the sternigenitoanal sharp variations between each species, with the majority of varia- shield. Post-stigmatic trachea as in adult female. tion found in ITS1 and 5.8S and almost no difference in ITS2. This is Remarks. Populations of D. apodis constitute a distal clade with contrary to findings concerning most other mites that have been strong bootstrap values in all combined analyses and in mitochon- sampled (Navajas and Fenton, 2000; Cruickshank, 2002), but sim- drial single gene analyses (Table 3, Figs. 2, 3 and 5, Appendix D). ilar to patterns found in Tetranychus species (Navajas et al., Publication III L. Roy et al. / Molecular Phylogenetics and Evolution 50 (2009) 446–470 461
1998). Resulting Dermanyssus sequences were easy to align and 3. Anal plate D-shaped, dorsal shield without any conspicuous displayed several differences useful for molecular identification subapical shrinkage...... D. hirundinis at the species level. Differences clearly characterize our respective 30. Anal plate more elongate and more or less subrectangular, populations of D. carpathicus, D. hirundinis, D. longipes, D. hirsutus, usually dorsal shield with a subapical shrinkage and ending with D. quintus, D. apodis and clearly separate these species from each a quite rounded apex...... D. longipes another. Anyway, some few differences also separate some of the 4. Pronotal scutella usually separated from dorsal shield and far D. gallinae (especially the special lineage one L1) and D. longipes from it, anterior part of dorsal shield more or less concave, forming populations. a neckline between the two anterior pores, anterior pores usually The mitochondrial markers (COI and 16S rRNA) show much far from dorsal shield...... D. apodis higher levels of genetic differentiation between species than 40. Pronotal scutella in most cases touching dorsal shield, ante- ITS (hierarchical level 2; Fig. 4), and exhibit small amounts of rior part of dorsal shield more or less convex...... D. change between populations of a given specific entity (hierarchi- gallinae, including the special lineage one. cal level 1 of pairwise divergence in 16S and COI, Fig. 4). Addi- tionally, some populations with identical ITS sequences exhibit 4.4. Phylogenetic relationships between species several differences in their 16S sequences (Fig. 4). COI is by far the most variable of the three genes tested and provides many Within Dermanyssidae, Liponyssoides possesses similar chelic- changes between populations. Pairwise divergence percentages eral and cornicular characters, but has a hexagonal-shaped sternal between several populations of D. gallinae collected from Euro- shield. Most species of Liponyssoides are also found on mammals pean layer houses show that this marker is likely appropriate instead of birds and it is unknown whether they are a sister group for phylogeographic investigation concerning economically to Dermanyssus or originate from within the genus. Unfortunately, important species. Gene sequences of COI revealed it to be a per- no specimens were available for inclusion in this study. tinent marker for phylogeographic exploration at a low taxo- The results of the morphological analysis indicate that only two nomic level (between closely related species or even internal nodes provide a strong relative Bremer support (1.00). populations of the same species) and although it is a protein First, D. trochilinis is sister to the rest of tested Dermanyssus species. coding gene, pairwise divergence appears often sufficient, if not Secondly, the group Microdermanyssus + hirsutus-group appears excessive in some organisms (DeSalle et al., 2005), for obtaining monophyletic. Within the group Microdermanyssus + hirsutus- valuable phylogeographic information, even in some parthenoge- group, the relationships of the three species D. quintus, D. hirsutus netic species (Heethoff et al., 2007). and D. grochovskae, which correspond to Moss’ (1968) hirsutus- group in the subgenus Dermanyssus, are unresolved. Three of the 4.3. Feedback on primary hypothesis: a diagnostic key for four remaining species correspond to Moss’ Microdermanyssus identification of the gallinae-group species from France and D. rwandae is sister to these three. This species is one of the more recently described species, and has never been included in As stated previously, many characters traditionally used for a generic review of the group. Present results suggest it may be a species identification have shown to be highly variable even within member of the subgenus Microdermanyssus. D. carpathicus, also de- single individuals of the various species of the gallinae-group. To scribed after Moss’ last review, is resolved as a sister group to the date, there are no clear elements available for determining Microdermanyssus + hirsutus-group clade, but with rather weak whether these differences are real phenotypic plasticity or only support values. Our current morphological results coincide with pure variations, because the impact of environmental influence Moss’ idealized phylogeny with the exception of the hirsutus-group on these variations is very difficult to estimate. Thus, it appears placement, which clearly falls out within Moss’ Microdermanyssus. that some confusion may have occurred in some of the previous re- The subgenus Dermanyssus appears to be paraphyletic. cords likely due to these variations. A very recent example is found The total evidence and combined molecular only datasets offer in Brännström et al. (2008), who found differences in ITS1 between much more resolution in the gallinae-group than the morphologi- some D. gallinae from layer farms and some D. gallinae from wild cal analysis. These species appear more differentiated than ex- avifauna. Conspicuously, the ITS1 sequence found from wild birds pected on the single basis of morphology. Whereas the corresponds to our D. longipes. morphology-only topology results in a comb of gallinae-group spe- Most of these characters are no longer useful for distinguishing cies, with only Microdermanyssus + hirsutus-group forming a sup- species, however, in light of the phylogenetic results and closer ported clade based on relative Bremer support, the combined examination of characters, the following key has been generated tree shows a gallinae-group split into two different clades, one of for use in discriminating several species of the gallinae-group. them involving D. carpathicus, D. longipes and D. hirundinis (clade Thus, among leg chaetotactic characters, it turned out that lateral I), the second one a complex of D. gallinae lineages (clade B). The sides of some leg articles have very few intraspecific variations, position of D. apodis is unresolved in the MP analyses, but is sister in contrast to most of other sides. to the hirsutus-group + D. carpathicus + D. longipes + D. hirundinis 1. al1 seta of palp genu spine-like, two setae on femur I d longer (clade G) in the Bayesian analysis. As for the only two tested spe- than the three others, pl of genu II and III usually each with one cies of hirsutus-group, their mutual position is not strongly sup- seta, pronotal scutella fused or not to dorsal shield (often asym- ported in these analyses. According to Bayesian analysis, they metric arrangement), anal plate D-shaped, anterior part of dorsal would be considered close to the clade I. This would be rather con- shield more or less convex...... D. carpathicus gruent with morphological topology, with D. carpathicus in a sister 10. al1 seta of palp genu lanceolate, setae on femur I d without position to the hirsutus-group. But this is not supported by MP any conspicuous difference of length, other characters analysis. diverse...... 2 2. pl of genu II and III usually each with one seta, anal plate 4.5. Host specificity D-shaped or elongate and more or less subrectangular, pronotal scutella usually unfused to dorsal shield and far from it, anterior Molecular results strongly confirm very low host specificity part of dorsal shield more or less convex...... 3 among D. gallinae, which has been previously suggested (Zems- 20. pl of genu II and III usually each with two setae, anal plate kaya, 1971; Nosek and Lichard, 1962; Zeman and Jurík, 1981; Fen- D-shaped, other characters diverse...... 4 dˇa and Schniererová, 2004). Populations of this species were Publication III 462 L. Roy et al. / Molecular Phylogenetics and Evolution 50 (2009) 446–470 collected in this study from eight different orders of birds (Passer- two different bird families as well (Passer and Parus, respectively iformes, Coraciiformes, Piciformes, Galliformes, Ciconiiformes, Col- Passeridae and Paridae), the first being the genus of the type host. umbiformes, Apodiformes, Strigiformes), including domestic and D. apodis has been found only on one species A. apus. All these spe- wild birds. This leads to the conclusion that parasite transfer be- cies have been found only in wild avifauna. tween wild bird fauna and domestic fowl is not out of the question. As for lineages in the D. gallinae clade (B), not only have several Additionally, within it, the special lineage one appears also rather of them been found in ‘‘human managed environment” (cf. Fig. 3B), unspecific, having also been collected from three different orders but also several lineages group populations from disparate bird of birds (Columbiformes, Apodiformes, Strigiformes, Table 1). species (cf. above). Especially, the clade D that groups together On the other hand, many species appear more host specific, at populations from layer farms, canary breeding facilities, wild Euro- least in France, such as D. hirundinis (found only on Hirundinidae) pean Roller, and a Woodpecker with strong support values (inter- and D. carpathicus (found only in nests of two genera, Parus and nal clades not supported). Moreover, the clade E includes Phoenicurus) and D. longipes (found in nests of two genera, Parus populations found only in wild avifauna (D. urbica, Hirundinidae and Passer). Additionally, D. apodis was collected many times from and P. major, Paridae). two different places in France (ca 300 km apart), from more than Finally, special lineage one, the more basal lineage of D. gallinae, 50 nests of A. apus and numerous individuals caught for banding. has been found in pigeon breeding facilities, also in pigeon nests in This is the only host species it is known from for the moment. town, but never in layer farms. This lineage does not appear abso- Some of these results contradict the published literature. D. hir- lutely specific, as it has been isolated from two other bird groups, undinis has been recorded from roughly 40 different bird species, in in natura (a owl and a swift; cf. Table 1). But in these two cases, 9 bird orders (for review, Roy and Chauve, 2007): Passeriformes mites were not necessary infesting the inhabitant of nest (a single (19 previously recorded genera), Anseriformes, Apodiformes, mite in each case, isolated dead and dried from the two nests). Pi- Sturniformes, Strigiformes, Coraciiformes, Ciconiiformes, Columbi- geon are known to be concurrent with swift concerning nesting formes, Piciformes. In the present study, representatives from six place, especially into a town, which could explain the presence of of the above bird orders have been tested, but individuals belong- the single specimen GO8. ing to D. hirundinis have been found in only three species of Hirun- In short, host specificity may appear higher at the basis of the dinidae (Passeriformes; Delichon urbica and Hirundo rustica in gallinae clade, with special lineage one (clade F) mainly on pigeons, France, and Tachycineta bicolor in the USA), which is the type host with the intermediate clade E only found in wild passeriforms, and family. Moreover, it was present in 25% of analyzed hirundinid finally with remaining distal populations isolated from disparate nests in France. This suggests that D. hirundinis is more specific bird groups. And synanthropicity appears to be proper to this clade in France than expected from published data. As no faunistic inven- B, the second one, clade G in Fig. 3B, being only found in wild avi- tory of Dermanyssus species in France was available today, it is fauna. Intermediate ecological characteristics can maybe be seen in likely that ‘‘dermanyssofauna” (and host specificity) differs in Eur- D. apodis and in some cases D. gallinae special lineage one, both ope and in the USA. having urban hosts (swift and pigeons).
4.6. Evolutionary hypothesis for tested species of Dermanyssus 4.7. Conclusion
Tested species of Dermanyssus split into two clades in the total The morphology-based phylogenetic hypothesis presented evidence (B and G in Figs. 2B and 3B) and molecular—only analy- herein involves 20 of the 24 currently recognized species and a ses. No clear evolutionary hypothesis can be drawn on the basis new species. The monophyly of Dermanyssus could not be tested of morphology from this topology. Thus, several characters change due to the lack of Liponyssoides specimens and no assessment of their state at the basis of the clade grouping tested species of Der- relationships between the Microdermanyssus + hirsutus-group and manyssus, but all reverse in one clade or another more distally (K1, the gallinae-group has been possible. Anyway relationships of K17, K26, K27, K40, K43). These homoplasies obscure a clear view species of the gallinae-group within the genus are robustly exam- of morphological evolution. If we consider some ecological traits, ined. These results suggest that Microdermanyssus + hirsutus- some clues can be found. group contains species which are clearly distinguishable from No strict comparison with outgroups can be done, due to the one another solely on the basis of morphology, whereas the different ways of life: A. casalis and T. pyri are predatory mites remaining species (gallinae-group) are sometimes indistinguish- (and even A. casalis is not able to be occasionally parasitic; Lesna able from one another. Molecular data obtained from several pop- and Sabelis, pers. comm.). Only O. bacoti is a parasite, but is from ulations of the gallinae-group indicates that many of these an unrelated family of mites. indiscernible species are clearly distinct species: D. gallinae and All strongly supported lineages here involve species found on a D. hirundinis are molecularly clearly divergent species, and D. gal- narrow host spectrum, except for the clade of D. gallinae. Even linae contains several lineages. D. carpathicus is also a valid spe- within this clade, an evolution of host spectrum seems to be visi- cies, it is present in France and, by mapping morphological ble, although host spectrum is not to be considered phylogeneti- characters onto the molecular phylogeny, it appears there are cally. Indeed, observation of obtained topologies in correlation two diagnostic characters (K4 and K17) for the species (Fig. 3A). with our bird data did not allow considering that there are any A special lineage constituted by the subset of populations COL, coevolution events (A. Cibois, pers. comm.). This suggests that this JGC1, PI, LC may be a cryptic species closely related to D. gallinae. genus includes species with opportunistic habits. But phylogeny D. apodis is a new species to be linked to the gallinae-group, found can be correlated to some ecological traits, difficult to clearly de- on the black swift A. apus. fine, but inducing variations in adaptability to environment (farms Based upon the way the tested populations of the gallinae-group vs natura, transferability from one to another bird species, etc.). D. sorted out, host specificity of D. gallinae appears very low, found on hirundinis has been found in three species of Hirundinidae (D. urb- domestic birds as well as several orders of wild birds. On the other ica, H. rustica, T. bicolor). D. carpathicus was found in two different hand, D. hirundinis and D. carpathicus appear much more host spe- passeriform genera distributed in two different bird families cific, only found on a restricted set of hosts, at least in France. D. (Phoenicurus and Parus, respectively, Muscicapidae and Paridae), apodis has been found very often on A. apus, and only on this spe- as is the distribution found in the literature (Zeman, 1979). D. long- cies in France. The fauna of Dermanyssus seems to strongly differ ipes has been found in two genera of Passeriformes distributed in between USA and Europe. Publication III L. Roy et al. / Molecular Phylogenetics and Evolution 50 (2009) 446–470 463
Finally, two different clades were revealed within tested species (Istituto Sperimentale per la Zoologia Agraria, Firenze, Italia), Hallan of the gallinae-group, one of which seems to develop synanthropic- J. (Texas A&M University, College Station, USA), Makarova O. (Severt- ity and proliferative capacity, with most derived OTUs present in sov Institute of Ecology and Evolution, Russian Academy of Sciences, hen farms. This also seems to be correlated with an adaptation to Moscow, Russia), Dusbábek, F. (Academy of Sciences of the Czech a wider host spectrum, populations with similar sequences being Republic, Ceske Budejovice, Czech Republic), van Nieukerken E.J. found on various bird orders in distal position in the gallinae clade (National Museum of Natural History Naturalis, Leiden, The Nether- on molecular tree. lands), Sabelis, M. and Lesna, I. (University of Amsterdam, IBED, Amsterdam, The Netherlands), Takahashi, M. (Kawagoe Sogo Senior Acknowledgments High School, Kawagoe City, Japan), Ono Hirotsugu (National Science Museum, Tokyo, Japan), Paoletti B. (University of Teramo, Teramo, We are grateful for their loan of reference material, sampling of Italy), Fend’a P. (Comenius University, Slovakia), Heuch P.A. and fresh material, help, and advice to Baker A.S. and Beccaloni J., Gjevre A.G. (National Veterinary Institute of Norway, Oslo, Norway), (Natural History Museum, London, UK), Huber B. (Zoologisches Cencek T. (Pan´ stwowy Instytut Weterynaryjny, Puławy, Poland), De Forschungsinstitut and Museum A. Koenig, Bonn, Germany), Rojas M. (Universidad de Sevilla, Sevilla, Spain), Chabrol P. (veteri- Judson M. and Bain O. (Muséum National d’Histoire Naturelle, nary office, Bourg-en-Bresse, France), Duret L. (Université Lyon 1, Paris, France), Beaulieu F., King Wan Wu and Lindquist E.E. (Agri- Lyon, France), Gory G. (Museum d’Histoire naturelle, Nîmes, France), culture & Agri-Food Canada, Ottawa, Canada), Klompen H. (Ohio Caparros O., Brucy L., Carrier L., Bouvier J.C., Vincent-Martin N., State University, Museum of Biological Diversity, Columbus, Beauvallet Y., and Dehorter O. (Centre de Recherche sur la Biologie USA), Ochoa R. (Systematic Entomology Laboratory, USDA, Belts- et les Populations d’Oiseaux, Muséum National d’Histoire naturalle, ville, USA), Klimov P.B. and Oconnor B.M. (Museum of Zoology, Paris, France), Tavernier P. (Centre de Soins aux Oiseaux sauvages du University of Michigan, Ann Arbor, USA), Wauthy G. and Peeters Lyonnais, Francheville, France), Veau F. (CORA Ardèche, France), M. (Institut Royal des Sciences Naturelles de Belgique, Brussel, Bonnet S. and Guichard N. (LEGTA Saint-Genis-Laval, France), Belgium), Kilpinen O. (Danish Institute of Agricultural Sciences, Lallemand G. (Lycée des Mandailles, Châteauneuf-de-Galaure, Lyngby, Denmark), Kreiter S. (Supagro, Montpellier, France), Knee France), Rigaux M. (IUT A, Université Lyon1, France). We also thank W. (University of Alberta, Edmonton, Canada), Zeman P. (Institute three anonymous reviewers for valuable comments, which greatly of Parasitology, Ceske Budejovice, Czech Republic), Nannelli R. improved the manuscript.
Appendix A
List of taxa examined for morphological analysis. Note: Indicated individuals correspond to specimens considered here as references. Other specimens of the same species may have been examined. T = type material; NT = non-type material.
Species Loan from Specimens’ status Dermanyssus alaudae (Schrank, 1781) British museum of natural history (London, UK; neotype) and Belgian royal institute T+NT of natural sciences (Brussel, Belgium; NT, A. Fain’s collection) Dermanyssus americanus (Ewing, British museum of natural history (London, UK; NT) and National museum of natural T+NT 1922) history (Washington, DC, USA)(T) Dermanyssus antillarum (Dusbábek Institute of parasitology (Ceske Budejovice, Czech Republic) T and Cˇerny´ , 1971) Dermanyssus brevis (Ewing, 1936) National museum of natural history (Washington, DC, USA) T Dermanyssus carpathicus (Zeman, Institute of parasitology (Ceske Budejovice, Czech Republic)and P. Zeman (Czech T+NT 1979) Republic), specimens from several field collections of Parus major and Phoenicurus phoenicurus in France Dermanyssus chelidonis Oudemans, British museum of natural history (London, UK;), and Agriculture and Agri-food NT 1939 Canada (Ottawa, Canada) Dermanyssus faralloni Nelson and National museum of natural History (Washington, DC, USA) T Furman, 1967 Dermanyssus gallinae (De Geer, 1778) British museum of natural history (London, UK; neotype)—Museum Koenig (Bonn; T+NT Germany; NT)—Muséum National d0Histoire naturelle (Paris, France; NT)—field samples Dermanyssus gallinoides Moss, 1966 Agriculture and Agri-Food Canada (Ottawa, Canada) T Dermanyssus grochovskae Zemskaya, Severtsov institute of ecology and evolution, Russian academy of sciences, (Moscow, NT 1961 Russia; NT, but specimens identified by A. Zemskaya herself) Dermanyssus hirsutus Moss and National museum of natural history (Washington, DC, USA; T) T Radovsky, 1967 Dermanyssus hirundinis (Hermann, British museum of natural history (Bonn, Germany;neotype), Koenig museum T+NT 1804) (London, UK;NT), Belgian royal institute of natural sciences (Brussel, Belgium;NT, Fain’s collection), National museum of natural history Naturalis. (Leiden, The Netherlands; coll. Oudemans, NT, nr P.4632) Dermanyssus nipponensis Uchikawa National science museum of Tokyo (Tokyo, Japan; TNSMT-Ac 12495) T and Kitaoka, 1981 Dermanyssus prognephilus Ewing, British museum of natural history (London, UK;NT), National museum of natural T+NT 1933 history (Washington, D.C., USA) (T, AL000244) and Ohio State University (Columbus, USA;NT) Publication III 464 L. Roy et al. / Molecular Phylogenetics and Evolution 50 (2009) 446–470
Appendix A (continued) Species Loan from Specimens’ status Dermanyssus quintus Vitzthum, 1921 Agriculture and agri-food Canada (Ottawa, Canada; T), Museum of Zoology, NT University of Michigan (Ann Arbor, USA; NT), W. Knee (NI), Dermanyssus rwandae Fain, 1993 Belgian royal institute of natural sciences (Brussel, Belgium;T) T Dermanyssus transvaalensis Evans British museum of natural history (London, Uk;T) T and Till, 1962 Dermanyssus triscutatus Krantz, 1959 National museum of natural history (Washington, D.C., USA;T) and agriculture and T+NT agri-food Canada (Ottawa, Canada, NT) Dermanyssus trochilinis Moss, 1978 National museum of natural history (Washington, D.C., USA; T) T Dermanyssus longipes (Berlese and Slovak national museum in Bratislava (Bratislava, Slovakia; 2 slides labelled D. NT Trouessart, 1889) passerinus, from specimens collected from prof. Milan Mrciak) from Passer domesticus (NT) and specimens collected from a nest of Passer montanus near Avignon (France; NT) Dermanyssus apodis n. sp. Specimens from several field collections from A. apus in France by G. Gory and G. Lallemand (T+NT) Haemogamasus hirsutus Berlese, 1889 MNHN (3D7, 3E1-3E4, 3E6, 3E10) NT Ornithonyssus bacoti (Hirst, 1913) Specimens from a live lab strain in MNHN (O. Bain) NT Androlaelaps casalis (Berlese, 1887) Specimens from several field collections in France NT Typhlodromus pyri Scheuten, 1857 Specimens from mite culture in the lab of S. Kreiter (Supagro, Montpellier)
Appendix B way, there exists very short peritremes in some species (type specimens), which appear really different than others (D. cheli- List of morphological characters and states used in the analysis. donis, D. alaudae, D. rwandae...), being almost atrophied. Be- The main source of the following characters was either direct tween this state of characters and all others, a gap is visible. observations or the following publications: Moss (1966, 1968, That is the reason why we encoded it differently, with only 1978) and Evans and Till (1962). two characters states (short, ie less than twice the diameter 1. Lateral contours of palp coxae in ventral view-0-Straight-1- of stigmate and long, ie more than 4 times the diameter of Convex. stigmate). 2. Sternal shield shape: relative location of points c and d, with c 10. Humeral paired simple pores as large as setae bases, on dorsal medially located on anterior margin of sternal shield and d laterally lo- shield-0-absent-1-present. cated on anterior margin of sternal shield-0-c at the same level as d 11. Humeral paired large pores, about 4 times larger than setae (anterior margin rather straight)-1-c located above the line be- bases and containing a central conical prominence (on or off dorsal tween both points d (anterior margin quite curved). shield)-0-absent-1-present. 3. Sternal shield shape: ratio e/a with e = width at the largest point 12. Apico-opisthosomal setae width-0-similar with shape and and a = central height-0-e/a>3-1-e/a 6 2. width to other setae-1-much wider and more massive. 4. Shape of seta al1 of palp genu-0-spine-like-1-lanceolate. 13. Apico-opisthosomal setae arrangement-0-in a jumble-1-regu- 5. Shape of post-stigmatic trachea (a tube extending posteriorly larly aligned. from each stigmate)-0-one curved tube, around coxa IV-1-in two 14. Ventro-opisthosomal setae, located on areas laterad to anal separate pieces. shield-0-‘‘classical” number-1-neotrichy. 6. Principal pore on post-stigmatic trachea (a pore located on post- 15. Dorsal setae: comparison between central/peripheral setae of stigmatic trachea, which extends posteriorly from stigmata):-0-ab- dorsal shield (series j4-6 and z5 / j2, z2, z4 and s4)-0-no major differ- sent-1-present, large (ca. 3 setae base), a hole usually surrounded ence-1-length of central setae less than 1/3 length of peripheral by a large raised chitinous ring (something like a neck)-2-present, setae. small (diameter smaller than setae bases) and simple (withtout 16. Relative length of setae on dorsal side of genu I-0-all quite the any neck). same length-1-One apical and one basal setae much longer than 7. Intermediate pore on post-stigmatic trachea (a small pore lo- others (> length of genu). cated between principal pore and stigmata, close to stigmata):-0- 17. Relative length of setae on dorsal side of femur I-0-all quite the absent-1-present. same length-1-Two apical setae much longer than others (> length 8. Ultrastructure of dorsal shield-0-grooves absent-1-grooves of genu). present. 18. Mesonotal scutella (are considered here only sclerotized areas 9. Relative length of peritreme-0- <2 diameter of stigmate-1- which detour more than 3 grooves of soft integument)-0-absent-1- >4 diameter of stigmate. present. The traditional character (peritreme length in relation with 19. Ampoula near internal margin of coxa IV-0-fuzzy outlines and the coxa it reaches) is a character which appeared to us not sharp apex (as if it was some crumpled membrane)-1-roughly to be reliable in any case as such, because of its own nature(see rounded, quite sclerotized. below K9 in K definitions). It is something soft, and superficial. 20. Ultrastructure of leg segments’ cuticule-0-smooth-1-em- It forms a narrow groove inserted in the integument along the bossed with large circles (about 2-3 on each side of segments podosoma. The position and length of this element vary from longitudinally). one to another mite from the same strain and traditionally used 21. Third seta on anal shield-0-absent-1-present. character states constitute a continuum, which suggest this is 22. Proportions of anal shield-0-as wide as long, apically not valuable species specific character within Dermanyssus. Any- rounded, D-shaped-1-wider than long, with lateral angles more Publication III L. Roy et al. / Molecular Phylogenetics and Evolution 50 (2009) 446–470 465 or less prominent-2-D-shaped, close to state 0, but longer than 41. Pronotal scutella (sticky or not to dorsal shield, usually rasp- wide and apically subrectangular. berry-shaped with 3-4 ‘‘berries”)-0-present-1-Pronotal scutella 23. Central longitudinal pillar within anal opening-0-present-1- absent. absent. 42. Shape of stigmata-0-roughly rounded-1-dewdrop shaped. 24. Relative location of anal opening on shield-0-anterior-1- 43. Relative width of anterior setae in hypostomal parallelogram posterior. according to al setae of palp femur-0-base of anterior setae (basal 25. Relative location of the largest part of anal shield-0-anterior-1- part of seta, not the pit receiving base of seta) quite as wide as base central. of palp setae-1-3 or more. 26. Proportions of tibia II-0-longer than wide-1-wider than 44. Outlines and shape of epigynial shield-0-Irregular contours, large-2-as long as wide. with a rather tapering apex-1-Sharp contours, with a rounded 27. Proportions of genu II-0-longer than wide-1-wider than apex, following a rounded narrowing. large-2-as long as wide. 45. Ampoula near internal margin of coxa IV-0-Conspicuously vis- 28. Morphometric comparison of oviporal flap (OF) and epigynial ible, with a narrowing at the base-1-Non clearly discernable or shield (ES): ratio length of OF/total length OF + ES-0-61/3-1-ca. = ½. simply a filiform element, such as a slight extension of poststig- 29. pv seta of palp trochanter-0-present, as narrow as next setae- matic trachea. 1-present, massive compared to next setae (large base, appears 46. Cornicules-0-Heavily sclerotized-1-Membranous. full)-2-absent. 30. al seta of trochanter I-0-present, as narrow as next setae-1- present, massive compared to next setae (large base, appears Appendix C full)-2-absent. 31. Anterior pair of setae within hypostomal parallelogram-0- Matrix of 24 taxa and 46 morphological characters used in the empty-1-filled with a clear substance. analysis. 32. Chelae-0-Mobile digit clearly distinguishable with an optical microscope-1-Mobile digit reduced, undistinguishable with an optical microscope. 33. Shape of 2nd cheliceral segment section-0-as wide as or wider than palp genu-1-narrower than palp genu. 34. j3-0-absent-1-present. 35. J3 and J4-0-off shield-1-J3 and J4 on shield-2-J3 on and J4 off or on limit. 36. Dorsal shield-0-Rounded, apical contours fuzzy-1-Posteriorly subtruncate, with two rounded ‘‘angles”. 37. Relative length of dorsal shield according to podosoma-0-same length-1-Dorsal shield much longer than podosoma (extending posteriorly, far behind coxae IV). 38. Relative width of dorsal shield according to podosoma-0-same width-1-Dorsal shield less wide than podosoma (lateral margin of dorsal shield not running accross each coxa). 39. Proportions of tibia I-0-longer than wide-1-wider than large- 2-as long as wide. 40. Proportions of genu I-0-longer than wide-1-wider than large- 2-as long as wide. Publication III 466 L. Roy et al. / Molecular Phylogenetics and Evolution 50 (2009) 446–470
Appendix D
Single gene analyses (COI, 16S, ITS) using MP and Bayesian analyses. Publication III L. Roy et al. / Molecular Phylogenetics and Evolution 50 (2009) 446–470 467 Publication III 468 L. Roy et al. / Molecular Phylogenetics and Evolution 50 (2009) 446–470 Publication III L. Roy et al. / Molecular Phylogenetics and Evolution 50 (2009) 446–470 469
Appendix E
Results of the molecular ‘‘all-taxa” matrix, involving all tested taxa, including those with only one gene sequenced. Bayesian analysis from 5,000,000 generations using partitioned data and independent models of evolution for each partition. Numbers on nodes refer to Bayesian posterior probabilities.
References lineages of the oribatid mite Platynothrus peltifer (Acari, Oribatida). J. Evol. Biol. 20, 392–402. Huelsenbeck, J.P., Ronquist, F., 2001. MRBAYES: Bayesian inference of phylogeny. Akaike, H., 1974. A new look at the statistical model identification. IEEE Bioinformatics 17, 754–755. Transactions on Automatic Control 19, 716–723. Johnson, L.S., Albrecht, D.J., 1993. Effects of haematophagous ectoparasites on Berlese, A., Trouessart, E., 1889. Diagnoses d’acariens nouveaux ou peu connus. nestling House Wrens Troglodytes aedon: who pays the cost of parasitism? Bulletin de la Bibl. Sci. de l’Ouest 2e année, 2e partie 9, 121–143. Oikos 66, 255–262. Brännström, S., Morrison, D.A., Mattsson, J.G., Chirico, J., 2008. Genetic differences in Katoh, K., Kuma, K., Toh, H., Miyata, T., 2005. MAFFT version 5: improvement in internal transcribed spacer 1 between Dermanyssus gallinae from wild birds and accuracy of multiple sequence alignment. Nucleic Acids Res. 32, 511–518. domestic chickens. Med. Vet. Entomol. 22, 152–155. Knee, W., 2008. Five new species of Rhinonyssidae (Mesostigmata) and one new Carroll, H., Beckstead, W., O’Connor, T., Ebbert, M., Clement, M., Snell, Q., McClellan, species of Dermanyssus (Mesostigmata: Dermanyssidae) from birds of Alberta D., 2007. DNA reference alignment benchmarks based on tertiary structure of and Manitoba, Canada. J. Parasitol. 94, 348–374. encoded proteins. Bioinformatics 23, 2648–2649. Laumann, M., Norton, R.A., Weigmann, G., Scheu, S., Maraun, M., Heethoff, M., 2007. Clayton, D.H., Tompkins, D.M., 1995. Comparative effects of mites and lice on the Speciation in the parthenogenetic oribatid mite genus Tectocepheus (Acari, reproductive success of Rock Doves (Columba livia). Parasitology 110, 195–206. Oribatida) as indicated by molecular phylogeny. Pedobiologia 51, 111–122. Cruickshank, R.H., 2002. Molecular markers for the phylogenetics of mites and ticks. Lewis, P.O., 2001. A likelihood approach to estimating phylogeny from discrete Syst. Appl. Acarol. 7, 3–14. morphological character data. Syst. Biol. 50, 913–925. De Lillo, E., 2001. A modified method for Eriophyoid mite extraction (Acari: Monaghan, M.T., Balke, M., Gregory, T.R., Vogler, A.P., 2005. DNA-based species Eriophyoidea). Internat. J. Acarol. 27, 67–70. delineation in tropical beetles using mitochondrial and nuclear markers. Philos. De Rojas, M., Mora, M.D., Ubeda, J.M., Cutillas, C., Navajas, M., Guevara, D.C., 2001. Trans. R. Soc. B 360, 1925–1933. Phylogenetic Relationships in Rhinonyssid mites (Acari: Mesostigmata) Based Moore, W.S., 1995. Inferring phylogenies from mtDNA variation: mitochondrial- on Mitochondrial 16S rDNA sequences. Exp. Appl. Acarol. 25, 957–967. gene trees versus nuclear-gene trees. Evolution 49, 718–726. De Rojas, M., Mora, M.D., Ubeda, J.M., Cutillas, C., Navajas, M., Guevara, D.C., 2002. Moss, W.W., 1966. Dermanyssus gallinoides n. sp. (Mesostigmata: Laelapoidea: Phylogenetic Relationships in Rhinonyssid mites (Acari: Mesostigmata) Based Dermanyssidae) an Acarine Parasite of Woodpeckers in Western North America. on ribosomal DNA sequences: insights for the discrimination of closely related Can. Entomol. 98, 635–638. species. Parasitol. Res. 88, 675–681. Moss, W.W., 1968. An Illustrated Key to the Species of the Acarine Genus DeSalle, R., Egan, M.G., Siddall, M., 2005. The unholy trinity: taxonomy, species Dermanyssus (Mesostigmata: Laelapoidea: Dermanyssidae). J. Med. Entomol. delimitation and DNA barcoding. Philos. Trans. R. Soc. Biol. Sci. 360, 1905–1916. 1, 67–84. Evans, G.O., 1963. Some observations on the chaetotaxy of the pedipalps in the Moss, W.W., 1978. The mite genus Dermanyssus: a survey, with description of Mesostigmata (Acari). Ann. Mag. Nat. Hist. 13, 513–527. Dermanyssus trochilinis, n. sp., and a revised key to the species (Acari: Evans, G.O., Till, W.M., 1962. The Genus Dermanyssus De Geer (Acari: Mesostigmata: Dermanyssidae). J. Med. Entomol. 14, 627–640. Mesostigmata). Ann. Mag. Nat. Hist. 13, 273–293. Moss, W.W., Camin, J.H., 1970. Nest parasitism, productivity and clutch size in Fendˇa, P., Schniererová, E., 2004. Mites (Acarina: Mesostigmata) in the nest of Purple Martins. Science 168, 1000–1003. Acrocephalus spp. and in neighbouring reeds. Biologia 59, 41–47. Navajas, M., Fenton, B., 2000. The application of molecular markers in the study of Goloboff, P.A., Farris, J.S., 2001. Methods for quick consensus estimation. Cladistics diversity in Acarology: a review. Exp. Appl. Acarol. 24, 751–774. 17, S26–S34. Navajas, M., Lagnel, J., Gutierrez, J., Boursot, P., 1998. Species-wide homogeneity of Goloboff, P.A., Farris, J.S., Nixon, K., 2008. TNT: Tree Analysis Using New Technology. nuclear ribosomal ITS2 sequences in the spider mite Tetranychus urticae Willi Hennig society ed. Available from:
Nosek, J., Lichard, M., 1962. Beitrag zur Kenntnis der Vogelnestfauna. Entomologické Samadi, S., Barberousse, A., 2006. The tree, the network, and the species. Biol. J. Linn. problémy (Bratislava) 2, 29–51. Soc. 89, 509–521. Nylander, J.A.A., 2004. MrModeltest v2. Program distributed by the author. Shaw, K.L., 2002. Conflict between nuclear and mitochondrial DNA phylogenies of Evolutionary Biology Centre, Uppsala University. a recent species radiation: what mt DNA reveals and conceals about modes Pacejka, A.J., Santana, E., Harper, R.G., Thompson, C.F., 1996. House Wrens of speciation in Hawaiian crickets. Proc. Natl. Acad. Sci. USA 99, 16122– Troglodytes aedon and nest-dwelling ectoparasites: mite population growth 16127. and feeding patterns. J. Avian Biol. 27, 273–278. Springer, M.S., DeBry, R.W., Douady, C., Amrine, H.M., Madsen, O., de Jong, W.W., Pacejka, J., Gratton, C.M., Thompson, C.F., 1998. Do potentially virulent mites affect Stanhope, M.J., 2001. Mitochondrial versus nuclear gene sequences in deep- house wren reproductive success?—Troglodytes aedon. Ecology 5, 79. level mammalian phylogeny reconstruction. Mol. Biol. Evol. 18, 132–143. Peterson, A. Version 8.07 (2007.08.25). ‘Zoonomen Nomenclatural data’ Available Swofford, D.L., 2001. PAUP*-Phylogenetic Analysis Using Parsimony (*and other from:
94
5 Ecologie comparée des cinq espèces françaises du genre Dermanyssus Cinq espèces ont été rencontrées au cours de la présente étude dans les nids d’oiseaux collectés en France, ou directement sur des oiseaux capturés en France : D. gallinae, D. carpathicus, D. longipes, D. hirundinis, D. apodis. Chacune de ces espèces a été rencontrée de manière récurrente dans les échantillons analysés et indépendamment de la provenance géographique. 5.1 Spécificité d’hôte chez cinq espèces du genre Dermanyssus : publication IV
a - Présentation Une fois les espèces redéfinies dans le groupe gallinae, le problème de la spécificité d’hôte peut enfin être abordé. Les espèces du groupe gallinae sont-elles vraiment toutes très généralistes ? Ou bien leur morphologie floue a-t-elle réellement introduit des erreurs dans l’inventaire, comme le craignait Moss (1978) ?
5.1.a.1 Objectifs L’objectif principal de la quatrième publication était d’obtenir un aperçu solide de la répartition des espèces du groupe gallinae communes en France en fonction des groupes d’oiseaux et d’en tirer des patrons de variabilité éventuels en lien avec les informations d’ordre pratique et écologique liées à nos échantillons.
5.1.a.2 Matériel et méthodes Une estimation du spectre d’hôtes par espèce a été réalisée, par le biais d’un échantillonnage large et d’une identification systématique des acariens du genre Dermanyssus isolés, au niveau spécifique. L’utilisation sur un nombre accru d’isolats de l’un des marqueurs moléculaires utilisés pour la redéfinition des espèces (mt-Co1) a permis d’établir un scénario de l’évolution du spectre d’hôtes au sein des espèces testées sur la base d’une reconstruction phylogénétique. Un complément d’information a été recueilli par le biais de petites expérimentations en laboratoire (transfert artificiel d’un hôte à l’autre et suivi du développement individuel). Cela a contribué à mettre en évidence certains patterns évolutifs quant à l’écologie de cinq espèces communes en France. Quelques essais de transfert d’un hôte à l’autre au laboratoire, ainsi que des observations à partir d’informations de terrain, ont permis d’approfondir quelque peu la réflexion sur les patterns écologiques révélés.
5.1.a.3 Principaux résultats (1) Les espèces du groupe gallinae ainsi que les deux espèces testées du groupe hirsutus se différencient en deux clades principaux sur la base de la mt-Co1. L’un de ces deux clades réunit des espèces rencontrées dans l’avifaune sauvage exclusivement, le second, constitué des différentes lignées composant D. gallinae, partagé entre avifaune sauvage et élevages (synanthropicité). Ces deux clades montrent par ailleurs deux niveaux de spécificité d’hôte (les spécialistes ne parasitent que l’avifaune sauvage – cf. Fig. 6).
94 95
Hôtes recensés T pyri O bacoti A casalis 1.00 GO54 MAR Apodiformes GO58a D. sp. GO36 D hirsutus D quintus LC23E 0.99 JBO59 Piciformes 1.00 LR20A 1.00 Ecop1 EcopO6-5 Ecop3a Ecop06 9a 0.98 RQ D. carpathicus 1.00 RQ_Mes1 Veol JMC10 LC10A RQ_Mes3 1.00 PAS ENVL08 3 ENVL08 8 ENVL08 1 D. longipes Passeriformes 0.95 HIR6b HIR1 ADhirun 0.99 1.00 TROAED 1.00 PARATR TACBIC CB3 CB5 D. hirundinis HIR6a CB4 0.99 OC CHOV HR 1.00 LB07 4 JBO517 LB18 0.74 LC 1.00 COL JGC PI CANIM Passeriformes 8008 0.85 F01 5013 Percno 8006a Columbiformes 8003b1 8002b 8009 Piciformes F22AR 0.70 F86 F50S ROL1 D. gallinae Strigiformes CANIT JB046 JBO27 Ciconiiformes 8007 Fa2 PO1 PO2 Coraciiformes 8004 Avifaune 8012 8006b Sturniformes Fa1 sauvage Chab DR SK Apodiformes Wood ROL2 IL302p Galliformes IL213 IL227 IL202 0.1
Figure 6. Reconstruction phylogénétique avec indication d’hôte et de milieu d’échantillonnage intégrant 73 isolats du groupe gallinae sur la base d’une région de la mt-Co1, analyse bayésienne, MrBayes v3.1.2, modèle d’évolution GTR+*+i pour 5 000 000 générations.
(2) Le spectre d’hôtes est fortement dépendant des habitudes de l’oiseau hôte, avec transfert possible d’un hôte principal à un autre en cas de partage de l’emplacement du nid chez certains oiseaux (cas des oiseaux vivant en colonies et nichant dans - ou reconstruisant un nid par-dessus - des nids de la colonie, mais pas nécessairement le même chaque année). (3) L’écosystème du nid (ou de la litière) semble comporter des paramètres essentiels au succès de l’adaptation de l’acarien et du développement de ses populations : la composition de l’arthropodofaune du nid (différentes guildes, telles celle des prédateurs, celle des détritivores, celle des parasites, en proportion variable et à la diversité variable en fonction de l’espèce d’oiseau, du site de nidification, …), des facteurs physico-chimiques (température, hygrométrie, taux atmosphérique d’ammoniac…), les habitudes hygiéniques de l’oiseaux (rejet ou non des sacs fécaux des poussins hors du nid, …), la disponibilité de l’hôte (tout au long de l’année, ou seulement durant la période de nidification), la présence de pesticides ou non, etc. sont autant de paramètres qui peuvent avoir une influence sur le développement des microprédateurs. Or ils sont par excellence très différents entre habitats sauvages et habitats anthropisés comme le sont les élevages. Le complexe de lignées composant D. gallinae
95 96
apparaît fortement opposé à l’ensemble des autres. Il semble intrinsèquement adapté aux conditions d’élevage. (4) L’oiseau est peut-être l’unique vecteur du microprédateur. Mais il n’est pas exclu que certains insectes volants puissent contribuer à la dissémination de l’acarien. Quoi qu’il en soit, il semblerait que l’acarien soit peu enclin à parcourir par lui-même des distances de plus de 2 mètres pour atteindre un hôte potentiel, même en cas de pénurie. (5) Les échanges entre volaille domestique et avifaune sauvage ne paraissent pas impossibles, mais ne sont pas encore prouvés.
b - Remarques sur la publication IV
5.1.b.1 Données complémentaires sur la spécificité d’hôte chez D. hirundinis en France D. hirundinis (13 isolats testés) s’est avéré spécifique aux Hirundinidae en France (10 isolats), alors que les 3 isolats provenant d’un même site des Etats-Unis ont été isolés chez trois familles différentes de Passériformes (dont les Hirundinidae). Outre la divergence génétique des séquences provenant de France et de celles provenant des Etats-Unis, l’explication de cette différence de spécificité développée dans la publication IV repose sur la différence d’écologie entre les espèces d’Hirundinidae concernées (l’hirondelle des arbres aux Etats-Unis est cavernicole et partage volontiers les nichoirs avec les deux autres familles recensées). Or, en France, si aucune des deux espèces d’Hirundinidae testées ne niche jamais dans des nichoirs, certains Passériformes d’autres familles profitent souvent de nids d’hirondelles déjà construits pour élever leur progéniture. Le moineau domestique (Passer domesticus) et le troglodyte mignon (Troglodytes troglodytes) sont coutumiers du fait. Jusqu’à la publication IV, aucun acarien apprtenant à D. hirundinis n’avait pu être isolé de nids de ces espèces, mais peu de ces nids avaient pu être testés. Depuis lors, un nid de troglodyte mignon a permis d’isoler 3 individus morts (LC083a, b, c) appartenant à notre lignée française de D. hirundinis. S’agit-il simplement d’individus de la saison précédente, qui se seraient en fait développés sur hirondelle ou ces individus ont-ils réellement parasité le troglodyte ? De plus, un individu mâle adulte (MG1) apparenté ou appartenant à D. hirundinis sur la base de sa séquence d'ARNr 16S (94-96% identité) et de sa séquence de mt-Co1 (96-99% identité) a été isolé dans la litière d’un jeune moineau vivant trouvé errant et hébergé dans une maison (probablement tombé du nid). Une question du même ordre que ci-dessus se pose. Toutefois, le fait qu’il s’agisse d’un mâle – sexe qui ne se nourrit pas au stade adulte et très rarement rencontré sur l'hôte - laisse supposer qu’il est arrivé à un stade inférieur (deutonymphe) et a réalisé sa métamorphose dans la litière, après un repas de sang pris sur l’oiseau. Ces nouveautés sont discutées dans la publication V (p. 214). Cela confirmerait le fait que D. hirundinis n’est pas fondamentalement inféodé aux hirundinidae, mais qu’il ne peut être transféré à un autre oiseau que par partage direct du nid. Il est probable que les acariens de cette espèce ne transitent que très rarement sur l’hôte (ou sur un quelconque insecte), demeurent longtemps dans le matériau du nid et parasitent tout oiseau venant s’installer et séjourner dans ce même nid. Le mâle ci-dessus a d’ailleurs probablement été emporté par hasard par l’oiseau s’envolant au moment précis du repas. Toutefois, la faible prévalence apparente de D. hirundinis chez les moineaux, sur la base de nos échantillons de nids de moineaux du genre Passer, suggère que la durabilité du système n’est pas optimale entre cette espèce de Dermanyssus et ce genre d’oiseau.
96
PUBLICATION IV
Exp Appl Acarol (2009) 48:115–142 DOI 10.1007/s10493-008-9231-1
Molecular phylogenetic assessment of host range in five Dermanyssus species
L. Roy Æ A. P. G. Dowling Æ C. M. Chauve Æ I. Lesna Æ M. W. Sabelis Æ T. Buronfosse
Received: 10 October 2008 / Accepted: 16 December 2008 / Published online: 22 January 2009 Springer Science+Business Media B.V. 2009
Abstract Given that 14 out of the 25 currently described species of Dermanyssus Duge`s, 1834, are morphologically very close to each another, misidentifications may occur and are suspected in at least some records. One of these 14 species is the red fowl mite, D. gallinae (De Geer, 1778), a blood parasite of wild birds, but also a pest in the poultry industry. Using molecular phylogenetic tools we aimed to answer two questions concerning host specificity and synanthropicity: (1) is D. gallinae the only species infesting European layer farms?, and (2) can populations of D. gallinae move from wild to domestic birds and vice versa? Mitochondrial cytochrome oxidase I gene sequences were obtained from 73 Der- manyssus populations collected from nests of wild European birds and from poultry farms and these were analyzed using maximum parsimony and Bayesian inference. Mapping of the observed host range on the obtained topology and correlation with behavioural observations revealed that (1) host range is strongly dependent on some ecological parameters (e.g. nest hygiene, exposure to pesticides and predators), that (2) out of five species under test, synanthropic populations were found only in lineages of D. gallinae, and that (3) at least some haplotypes found in wild birds were very close to those found in association with domestic birds.
Keywords Dermanyssus Bird parasite mt-COI Host range Synanthropicity Host transfer
L. Roy (&) C. M. Chauve T. Buronfosse Universite´ de Lyon, Ecole Nationale Ve´te´rinaire de Lyon, Laboratoire de Parasitologie, Marcy l’Etoile, France e-mail: [email protected]
A. P. G. Dowling Department of Entomology, University of Arkansas, Fayetteville, AR, USA
I. Lesna M. W. Sabelis University of Amsterdam, IBED, Amsterdam, The Netherlands 123 PUBLICATION IV
116 Exp Appl Acarol (2009) 48:115–142
Introduction
Dermanyssus gallinae (De Geer, 1778) is a nidicolous mite that is well known as a pest in poultry farms. Apart from this species, there are 24 other species that are currently included in the genus Dermanyssus Duge`s 1834 (Roy and Chauve 2007; Knee 2008; Roy et al. 2008), but 14 of these species are morphologically very similar to D. gallinae, known as the poultry red mite. This may lead to identification problems and molecular tools are needed to answer questions concerning host specificity and synanthropicity: is D. gallinae the only species infesting European farms? Can populations of D. gallinae from wild birds and those from domestic birds undergo genetic exchange? Host–parasite relationships, and especially host specificity, are hard to define in many mite groups, because they are often opportunistic, and they may have multiple hosts during post-embryonic development. For example, the ticks have been divided into 6 different categories according to the relative width of their host range and according to the mono- or polyxeny during development (Hoogstraal and Aeschlimann 1982). A similar situation exists with respect to species in the genus Dermanyssus. Although these species seem to complete their whole life cycle on a single bird species, the amplitude of their host range varies between two groups of species. According to previous records, only a few species appeared to be specialists in a single bird family within Derma- nyssus, such as D. alaudae (Schrank, 1781) (Alaudidae only), D. quintus Vitzthum 1921 and D. hirsutus Moss and Radovsky 1967 (Picidae only) (Roy and Chauve 2007). On the other hand, most of the species in the Moss’ gallinae-group, are known for having a very large host spectrum, involving a variety of bird families, widely distributed in bird phylogeny. D. gallinae and D. hirundinis (Hermann, 1804) have been recorded in the literature, respectively, from 8 to 9 different bird orders, some of which are phyloge- netically very distant. For instance, D. gallinae can develop in some Galliformes as well as in some Passeriformes, the former being basally and the latter distally situated in the large clade of Neognathae according to the phylogenetic reconstruction of birds pro- posed by Livezey and Zusi (2007). Not to mention several mammalian recorded parasitized species. However, interpretation of data available to date is blurred by misidentifications resulting from confusion on morphological discrimination and host specificity is likely to slightly differ in some species in the gallinae-group (Roy et al. 2008). In France, 5 species are commonly found in wild avifauna, which all belong to Moss’ gallinae-group (Roy et al. 2008). D. gallinae, the Chicken Red Mite, seems to be the only species encountered in farms and breeding facilities. Present in more than 80% in layer farms in Europe, it is an important pest, inducing sanitary problems and financial losses. As D. gallinae is a haematophagous mite, a pest in poultry industry and potentially a vector of some pathogens (Valiente Moro et al. 2005, 2007), it is of practical interest to assess if there are genetic exchanges between populations hosted by wild avifauna and domestic fowl. The aim of the work described in this article is to assess host specificity and host range in five Dermanyssus species by using molecular tools of phylogenetic analysis. An inventory of Dermanyssus species collected from wild and domestic birds in France and the Netherlands is provided as a first step towards a more comprehensive analysis. Moreover, the exploration of host specificity using phylogenetic tools will be comple- mented by some bioassays.
123 PUBLICATION IV
Exp Appl Acarol (2009) 48:115–142 117
Materials and methods
Field sampling
For birds that re-use the nest, part of the nests was collected after the birds had left the nest. Otherwise whole nests were collected. Most of wild birds’ nests were collected in France, and a few were collected in The Netherlands and in the USA. In addition, hundreds of wild birds have been directly examined in France. Moreover, some mite populations were collected from layer hen farms (from France, Norway, Denmark, Poland, Belgium) and facilities for breeding canaries, pigeons and chickens (France, Italy, Spain). A list of DNA- tested mite populations is provided in the Appendix. A population corresponds to a group of Dermanyssus mites found from a single nest (or a single building in a farm or breeding facility). Note that there was never more than 1 species of Dermanyssus per nest. Note: We used the bird classification according to Peterson (2008), except for the blue tit, which we referred to as Parus caeruleus instead of Cyanistes caeruleus (L, 1758), in order to match common use in bird banders.
Restricted study areas
In addition, ten special areas with a rather small diameter (\3 km) allowed us testing several different nests used by a single bird community. These restricted study areas are described in Table 1 and indicated by the following acronyms: CB, Ecop, ENVL, HIR, JBO, LB, MOL, RQ in France, IL in The Netherlands, BMOC in the USA.
Nests’ analysis
Nests were analysed using a method described by De Lillo (2001) except that no sodium hypochlorite was added to the water solution to wash the stack of sieves and that the sieves had a somewhat different mesh width (top to bottom: 2500, 1400, 180, 100 lm).
Molecular analysis
DNA was extracted from individual mites following a protocol that preserves an intact cuticle for voucher preparation and microscopic observation. Of each sampled population, 2 or 3 mite individuals have been sequenced. A 700–800 bp amplicon of mt-COI gene was isolated by PCR, depending on primer pairs used (i.e. on concerned species, cf. Table 2 and Appendix for EMBL database accession numbers), and then sequenced. PCRs was performed in either a Biometra TGradient or a MWG AG Biotech Primus 96 plus thermal cycler in typical buffer containing 2 ll of template DNA, 2.5 units of Taq polymerase, 10 nmol of dNTPs, 20 pmol of each primer and 3.4 mM MgCl2. After an initial denaturation step (95 C) for 10 min, followed by 40 cycles of 20 s at 95 C (denaturation), 30 s at 52 C (hybridization), and 90 s at 72 C (extension). A final extension step was carried out for 10 min at 72 C. Several primers designed for ampli- fication of DNA from various species are listed in Table 2 and were choosen to perform PCRs under the same conditions. Negative and positive controls were run with each round of amplification. PCR products were checked by electrophoresis in a 1% agarose gel. Depending on the brightness of the band either additional PCRs were run on the original template or reamplifications of the
123 1 x plAao 20)48:115–142 (2009) Acarol Appl Exp 118 123 Table 1 Description of restricted study areas (cf. ‘‘Materials and methods’’) Site Location Site characteristics Host Number of Number of nests Remarks acronym nests tested with Dermanyssus
ENVL France, 69 10 nestboxes occupied by great tits Parus sp. (Paridae) 11 6 D. longipes in 1 nest of 6 tested in or blue tits on the Campus of the 2007; D. longipes in 3 nests of 5 National Veterinary School of tested in 2008 Lyon from 2007 to 2008 Ecop France, 42 Natural protected area Ecopole with Parus major (Paridae), 15 7 D. carpathicus found only from several dozens of nestboxes, Phoenicurus ochruros Parus sp.—In 2006 and 2008. sampled in automn 2005, summer (Muscicapidae) 2006 and winter 2008 RQ France, 42 House, small garage with small hen Parus major (Paridae), 55 D. carpathicus abundant in redstart house and girder with a redstart Phoenicurus ochruros nest in 2006, 2007and 2008. Also nest and two tit nests in a natural (Muscicapidae) found in a tit nest sampled in 2006. protected area at c. 950 m altitude D. gallinae in hen house (Pop. (sampled in 2006, 2007 and 2008) 8004) MOL France, 69 A small farm housing sheep, calves, Hirundo rustica 3 0 Only one specimen of D. gallinae rabbits, dogs and cats, and (Hirundinidae) (Pop. 8005) collected from hen chickens in four layer hen houses, house in spring-summer 2008 and with wood girders, housing (several liters of litter analyzed) many swallow nests HIR France, 85 House with 4 swallow nests in three Hirundo rustica 65 D. hirundinis collected in 2 different rooms (Hirundinidae) successive years (before nesting in winter 2007, after nesting in spring 2008) CB France, 01 Small old farm housing calves, cats Hirundo rustica 8 3 Spring-summer 2008. D. hirundinis and also free-renge chickens (Hirundinidae) present in swallow nests. Many D. according to the production gallinae in wooden chicken cages procedure called ‘‘AOC Poulets de (Pop. 8012)
Bresse’’, and many swallow nests P UBLICATION I V x plAao 20)4:1–4 119 48:115–142 (2009) Acarol Appl Exp Table 1 continued
Site Location Site characteristics Host Number of Number of nests Remarks acronym nests tested with Dermanyssus
JBO France, 13 15 plots (maintained with chemical, Parus major (Paridae) 62 5 Summer 2007. One, three and no organic or alternative methods) of nests with D. gallinae in resp. apple/pear orchards with one alternative control plot (Pop. nestbox with great tits every 50 JBO27), organic plot (Pop. JBO51, meters in each row of trees JBO46, JBO56) and chemical control plot. One nest with D. carpathicus in organic plot (Pop. JBO59). IL The Netherlands, 33 nestboxes occupied by European Sturnus vulgaris (Sturnidae) 33 32 Summer 2007. D. gallinae abundant Groningen starlings (populations IL213, in many nests. IL227, IL302, IL202) BMOC USA, Michigan 4 nests with different birds from the Tachycineta bicolor 4 3 Autumn 2007. D. hirundinis rather campus of a school in Michigan (Hirundinidae), Parus abundant. atricapillus (Paridae), Troglodytes aedon (Certhiidae) LB France, 45 Several swallow nests from a goat Delichon urbica, Hirundo 10 9 Summer 2006 and 2007. D. gallinae farm rustica (Hirundinidae) present in some nests, D. hirundinis in some others. PUBLICATION IV 123 PUBLICATION IV
120 Exp Appl Acarol (2009) 48:115–142
Table 2 Primer sequences Primer sense Primer name Sequence 50-30
Forward CO1RQF1 GAAAGAGGAACAGGAACAGG CO1LCF GAAAGAGGAGCAGGCACTGG COF1bis CTGCACCTGACATGGCTTTCCCAC CO1F4 CACCTGACATGGCTTTCCCACGAT RhipiCOIF CGAATAAATAATATAAGATTTTGA SKPOFa2diagF1 CTTTTTAGATCTTTAATTGAAA Reverse COIGOR GTTGGAATtGCAATAAT RQ-COI-R CCAGTAATACCTCCAATTGTAAAT ObCOIF-rev GTGGGAATHGCAATAAT TyphloCOIR GCTAATCAAGAAAAAATTTTAAT
Primer pairs used in present study for the amplification of mt-COI according to species indicated into brackets: CO1RQF1 ? RQ-COI-R (D. carpathicus, D. hirundinis, D. longipes), SKPOFa2di- agF1 ? RQCOIR, CO1F4 ? RQCOIR, COF1bis ? RQ-COIR, COF1bis ? ObCOIF-rev, CO1LCF ? RQ- COI-R (D. gallinae), RhipiCOIF ? TyphloCOIR (outgroups), CO1LCF ? COIGOR (D. apodis) original PCR product were performed. PCR products were sequenced by Genoscreen (France, Lille) using a 96-capillary sequencer ABI3730XL.
Phylogenetic reconstruction
Sequence data was aligned using MAFFT (Katoh et al. 2005) with the L-INS-i iterative refinement option on the MAFFT server at http://align.bmr.kyushu-u.ac.jp/mafft/ online/server/. MAFFT with the L-INS-i option has shown to be the most accurate and consistent method for the alignment of sequences (Wilm et al. 2006; Carroll et al. 2007). The alignment of 558 bp from cytochrome oxidase I (mt-COI) was analysed under several optimality criteria: (1) Parsimony using PAUP* 4.0b10 (Swofford 2001) to build tree(s) with TBR branch swapping and 10,000 random additions saving all most parsi- monious trees, (2) Bayesian inference using the computer program MrBayes v3.1.2 (Huelsenbeck and Ronquist 2001; Ronquist and Huelsenbeck 2003) with the GTR ? C?i model of evolution chosen in the computer program MrModeltest (Nylander 2004) that was run for 5,000,000 generations. Burnin was determined based on stationarity being reached. The clade notation taxon? indicates the clade including the taxon at its base and all subsequent taxa included in the same clade (De Souza Amorim 1982).
Bioassays
In order to get some data complementary to phylogenetic information, some bio-assays were performed. Eight different bioassays have been performed in order to compare the behaviour of four Dermanyssus species. Note that most of these bioassays are rather tentative (method not completely validated), provide limited information, and are often not mutually compa- rable. However, some of the data obtained add interesting elements to the discussion. The first two (comp1 and 2) involved a permanent or intermittent contact of mites with one canary, during a period of several weeks (cf. Table 3). A single canary was placed in a plastic box, provided with several apertures on the bottom and sides covered with a filter tissue for aeration, and with an eating and a drinking trough that can be filled from the exterior of the box. This enclosure was placed into a large bowl filled with water and a drop 123 x plAao 20)4:1–4 121 48:115–142 (2009) Acarol Appl Exp
Table 3 Some data on development of 3 Dermanyssus species on canaries, obtained from long-time bioassays (referred to in text as comp1 and comp2) Bio-assay Host Mite species Mite population Total number Duration of Total number of of mites tested mite-bird contact living mites collected
comp1 a Canary D. gallinae SK 20 98 days 205 (permanent contact) b Canary D. longipes PAS 40 98 days 0 c Canary D. longipes PAS [200 (nest residu) [100 days 12 comp2 Canary D. carpathicus ? RQ ? SK ? Fa1 [200 (nest ca. 24 days [1000 D.gal.; 8 D.car. (intermittent contact: during 12 h every D. gallinae residu ? mites) 2 weeks, all along 1 year) Bioassay comp1 is split into three lines (a, b and c), which correspond to three boxes, each containing one canary, and which have been handled simultaneously and in the same conditions. Information about populations is to be found in the appendix. Mite population’s codes correspond to population codes in Appendix P 123 UBLICATION I V PUBLICATION IV
122 Exp Appl Acarol (2009) 48:115–142 of tension-active agent. Water and food were added regularly and the canary was allowed to move, feed and drink ad libitum. Mites introduced into the enclosure cannot escape and no mites from outside can get in. Moreover, the escape of mites from the enclosure into the surrounding water was regularly checked, which represents less than 10% of final count. At the end of the period, the litter and plastic box were rinsed and treated following the method of De Lillo (2001). Mites were observed and counted using a stereoscopic mag- nifying glass. The 5 other bioassays (comp3 to 8) consisted of short periods of contact between the mites and bird (several hours at a time) in glass containers (cf. Table 4). One small bird (canary, hen chick, duck chick) was placed within the container, which was placed into a bowl filled with water and a drop of tension-active agent. Next, mites were added (an aggregate consisting of an unknown number of individuals, or–in some cases–of a known number of individuals) and the whole device was kept in darkness (incubator at 25 C). In case of known number of handled mite individuals, mites were collected at lab using a vacuum pump and 10 lL filter tips (tips’ filters are permeable to air, but retain mites). Tips were closed with some parafilm for storage and broken just before being introduced in the glass container with the canary at the beginning of the bioassay. After several hours, birds were removed, devices were examined and engorged mites were isolated in ELISA mi- croplates, covered with extended parafilm (one small cut above each well, insertion of a single mite using a fine wet brush, obturation of cuts using a small piece of extended parafilm). Cues that provide information on developmental progress (exuvia, eggs) were recorded using a stereoscopic magnifying glass at d ? 4 and noted down as follows: 1 exuvia (protonymph, deutonymph) or 1 laying (1 to several eggs) (adult female) in a well indicates that the isolated individual has developed.
Results
Inventory of Dermanyssus species from wild and domestic bird
Overall, 27 bird species distributed across nine different bird orders were examined. Among wild avifauna, 334 nests of wild birds, representing 25 bird species, distributed across seven bird orders and 31 families, have been analysed (Table 5). Note that the goal here is not to get an overview of prevalence and that this study is not an epidemiological one. Here, we report the results of our explorations on host specificity, based on simple records on a batch of samples obtained from various bird species and places. That is the reason why even some bird groups with only few nests examined are noted. Anyway, the focus will be on bird groups with a significant number of samples analysed (Fig. 1). Moreover, hundreds of wild birds have been examined including Picidae, Alaudidae, Coraciidae, Paridae (adding one more bird species to our study: Coracias garrulus L., 1758). Additionally, several dozens of mites have been collected from layer hen farms (hence one more species included in our study: Gallus gallus L., 1758) and some breeding facilities (chicken, pigeons, canaries, included in Fringillidae in Table 5). Dermanyssus was present in nests of 10 wild bird families (Table 5). Moreover, one additional bird family parasitized by Dermanyssus was found by direct on-host sampling (during bird banding and bird care activities): C. garrulus (Coraciidae: Coraciiform) (Populations ROL1 and ROL2 in Appendix). Overall, in France and in the Netherlands, 5 species of Dermanyssus have been isolated from nests and birds: D. gallinae including a special lineage, which may represent a cryptic 123 x plAao 20)4:1–4 123 48:115–142 (2009) Acarol Appl Exp Table 4 Engorgment and development of 3 Dermanyssus species compared using three host species and short-time bioassays (referred to in text as comp3 to comp8)
Assay Host Mite species Mite Total Duration of Number of % of live mites Total Number of % of mites Number of Remarks population number of contact mite- engorged that were number of development having individuals mites tested bird mites engorged upon mites cues (d ? 4) developed dead after (hours:minutes) collected collection recovered at d ? 4 d ? 0 (including (mites those not engorged engorged and and collected submerged alive) into water)
comp3 Duck chick D. gallinae SK unknown 12:00 120 –NE120 100 NE comp4 a Canary D. carpathicus RQ unknown 04:40 3 –NE003 All mites were dead when collected b Canary D. gallinae 8,010 unknown 04:40 51 –NE27 53 1 c Canary D. hirundinis HIR5 unknown 04:40 45 –NE15 33 1 comp5 a Hen chick D. carpathicus RQ 154 06:30 0 000–0 b Hen chick D. gallinae 8,010 154 06:30 83 54 97 49 59 7 c Canary D. carpathicus RQ 154 06:30 0 000–0 comp6 a Canary D. gallinae 8,010 154 11:45 29 19 86 24 83 2 b Hen chick D. gallinae 8,010 154 11:45 60 39 90 46 77 4 c Hen chick D. gallinae 8,010 154 11:45 30 19 61 10 33 14 comp7 a Canary D. gallinae 8,010 154 05:00 31 20 78 22 71 0 b Canary D. gallinae 8,010 154 05:00 22 14 70 10 45 3 c Hen chick D. gallinae 8,010 154 05:00 10 6804403 At least 77 individuals (N1) remaining within the filter
tip, likely due P 123
to diarrhea UBLICATION I produced by chick V 2 x plAao 20)48:115–142 (2009) Acarol Appl Exp 124 123 Table 4 continued
Assay Host Mite species Mite Total Duration of Number of % of live mites Total Number of % of mites Number of Remarks population number of contact mite- engorged that were number of development having individuals mites tested bird mites engorged upon mites cues (d ? 4) developed dead after (hours:minutes) collected collection recovered at d ? 4 d ? 0 (including (mites those not engorged engorged and and collected submerged alive) into water)
d Hen chick D. gallinae 8,010 154 05:00 48 31 147 27 56 2 At least 77 individuals (N1) remaining within the filter tip, likely due to diarrhea produced by chick comp8 a Canary D. carpathicus RQ [150 12:00 8 \5% 33 ––– The 8 engorged individuals found dead in liquid canary’s droppings. b Canary D. gallinae 8,010 [150 12:00 82 \55% 95 –––
Subdivisions in comp4 to comp8 (a, b, c, d) correspond to the different glass containers involved in the bioassay, containing each one small bird, and which have been handled simultaneously and in the same conditions. Recorded development cues are exuviae and eggs. Information about populations is to be found in the appendix NE not estimated PUBLICATION IV x plAao 20)4:1–4 125 48:115–142 (2009) Acarol Appl Exp Table 5 Number of nests analysed per bird family and occurrence of genus Dermanyssus based and on present field data and on literature
Bird order Bird family Nests Mite Wild bird analysed populations individuals in from farms on which present or mite D. brevis D. apodis D. apodis D. quintus
study breeding populations D. alaudae D. gallinae D. hirsutus D. longipes D. rwandae D. faralloni D. trochilinis facilities have been D. chelidonis D. triscutatus D. wutaiensis D. hirundinis D. passerinus D. antillarum D. gallinoides D. carpathicus D. americanus D. brevirivulus directly D. nipponensis D. grochovskae D. prognephilus
found D. transvaalensis
Apodiform Apodidae 52 7 F L L L Trochilidae L Ciconiiform Accipitridae 3 L F L Falconidae 1 Alcidae L Ardeidae 2 Ciconiidae 1 Hydrobatidae L Columbiform Columbidae 9 2 1 L L Coraciidae 2 F Coraciiform Meropidae L L Galliform Phasianidae 18 M Gruiform Gruidae 1 Passeriform Alaudidae 14 L L L Certhiidae 1 M Cinclidae 1 Corvidae 1 L Fringillidae 4 3 L L M L L Hirundinidae 39 L L M L L L Laniidae L Motacillidae L Muscicapidae 23 M L L L L L Paridae 124 M L M M M L Passeridae 30 L L L L M L L Sittidae L L L Sylviidae 1 L L Vireonidae L Piciform Picidae 1 1 M L L L L L L L L Strigidae 2 L L L Strigiform P 123 Tytonidae 1 F
Sturniform Sturnidae 33 L M L L L UBLICATION I
Dermanyssus V PUBLICATION IV
126 Exp Appl Acarol (2009) 48:115–142
120
D. gallinae 5% D. gallinae 16% D. gallinae 8% D. gallinae 100% 100 D. «apodis » 63% D. hirundinis 52% D. carpathicus 40% 80 Unidentified 32% Unidentified 32% D. longipes 16% Unidentified 36% 60
40 % occurence
20
0 Apus apus Alaudidae Hirundinidae Parus spp. Sturnus vulgaris
Fig. 1 Percentage of occurrence of Dermanyssus in nests of the five bird groups under test. Above each column the percentage of identified species among Dermanyssus individuals is given
species (here referred to as L1), D. hirundinis, D. carpathicus (Zeman 1979), D. longipes (Berlese and Trouessart 1889) and D. apodis Roy et al. 2008). The number of nests analysed per bird group varied considerably, but a substantial number of nests were available in five bird groups and these represented together ca. 80% of all nests analysed. These groups are scrutinized below.
Apus apus, the black swift
Of the 52 nests analysed, 79% contained Dermanyssus individuals. Of these Dermanyssus, 32% were not identifiable at the species level because of their poor preservation condition. Of the remainder, 63% was identified as D. apodis and 5% as D. gallinae. In addition to Dermanyssus spp. collected from nest material, 18 individuals belonging to D. apodis were collected directly from the host (on several bird individuals), seven of which were on chicks in the nest and 11 on adults or on a fledged young (found far from nest).
Sturnus vulgaris, the European starling
Only one of the 33 nests analysed did not harbour any Dermanyssus individuals. Of the 32 others, four populations have been sequenced (mt-COI), which all appeared to belong to D. gallinae. Moreover, haplotypes found in these populations appeared to be very close to each other.
Parus major and P. caeruleus, the great tit and the blue tit
On the whole, 120 nests have been tested in the two species of tits, 62 of which originated from apple and pear orchards. Of all these nests, 21% contained some Dermanyssus individuals, but in orchards, 8% provided some Dermanyssus individuals, versus 34.5% in natura. Moreover, almost all individuals collected from orchards were identified as D. gallinae, whereas mites from nests sampled in natura were identified as D. carpathicus or D. longipes. Moreover, several hundreds of bird individuals have been examined, without finding any individual of Dermanyssus spp.
123 PUBLICATION IV
Exp Appl Acarol (2009) 48:115–142 127
Alaudidae, the larks
Nests of three species of Alaudidae have been examined, one of these species studied most extensively was Melanocorypha calandra L., 1766. No Dermanyssus were found in any of the nests. Moreover, no more Dermanyssus have been found on any of the hundreds of bird individuals (mainly Alauda arvensis L. 1758) examined from two different regions in France around 800 km apart (Droˆme, department 26, Pas de Calais, department 62). In particular, 50 individuals caught by a hunter in Droˆme were closely examined, then sub- merged in water and analysed using the De Lillo’s method, but there was no recovery of any Dermanyssus individual.
Hirundinidae, the swallows
Nests were obtained from two species of hirundinids: Delichon urbicum L., 1758 and Hirundo rustica L., 1758. Of the 42 nests analysed, 58% provided Dermanyssus individ- uals (mostly D. hirundinis). Moreover, the guild of bird parasites appeared to be more diverse, with several groups in addition to Dermanyssus, such as fleas (Insecta: Sipho- naptera), chewing lice (Insecta: Mallophaga) and an individual belonging to Myonyssus sp. (Mesostigmata: Laelapidae).
Phylogenetic analysis
The data matrix consisted of 558 characters from a coding region of cytochrome oxidase subunit 1 (COI), of which 216 were parsimony-informative. The 78 included taxa, cor- responding to mite populations, include 3 distant outgoups, 2 species of the hirsutus-group, and 73 populations of gallinae-group (groups classified according to Moss (1978)). Note that different haplotypes within a single population was detected in only two populations (RQ-Mes and 8006, 2 haplotypes of a single species detected in each), which come, respectively, from one nest and one farm. Parsimony analysis recovers 1000 most parsi- monious trees (L = 775, CI = 0.5316, RI = 0.8879, Fig. 2), with 55 distinct haplotypes of Dermanyssus (53 of gallinae-group). Bayesian analysis resulted in a topology similar to MP, slightly more resolved concerning internal relations of the D. apodis? clade (Fig. 3). The same groupings appear in analyses using a subset of taxa and a combined matrix concatenating a region of rRNA 16S, a region from rRNA 18S to 28S, including ITS1, 5.8S and ITS2, as well as the COI region (Roy et al. 2008).
Populations sharing haplotypes
Populations, as defined above, come from different nests or different farms or breeding facilities, from different places. Most of mites obtained come from France (cf. Fig. 4 in Appendix), some additional samples come from Belgium, Denmark, The Netherlands, Spain, Italy, Norway, Poland and the USA (cf. Appendix). Each population is at least 3 km apart from each another, except in the cases of restricted study areas (Table 1) in which some nests stood several hundreds meters apart from each other. Some haplotypes are to be noticed, as these were found in several populations of D. hirundinis and D. gallinae from single restricted study areas. For example, a single haplotype occurred in populations from areas BMOC, CB and IL, (except in two popu- lations, which had haplotypes differing in only one (IL 227) or two (CB3) nucleotide 123 PUBLICATION IV
128 Exp Appl Acarol (2009) 48:115–142
T pyri O bacoti A casalis D hirsutus D quintus 100 GO54 MAR GO58a GO36 D. apodis LC10A 67 Veol JMC10 62 Mes3 100 80 RQ Mes1 LC23E D. carpathicus Ecop1 LR20A 99 JBO59 5 Ecop3a Ecop06-9a 100 PAS ENVL08-3 ENVL08-8 D. longipes ENVL08-1 69 HIR6b HIR1 ADhirun 100 100 TROAED PARATR TACBIC 100 CB3 99 CB5 D. hirundinis HIR6a 81 CB4 OC CHOV HR L2 100 LB07-4 JBO517 67 LB18 78 CANIM 100 8008 LC 91 COL L1 JGC PI F01-5013 Percno 98 CANIT ROL1 98 8006a A 8003b1 8002b 8009 F22AR F86 D. gallinae 100 F50S JB046 79 JBO27 8007 Fa2 PO1 PO2 B 91 8004 8012 8006b SK Fa1 92 95 Woodp ROL2 65 Chab SB IL302 78 IL213 IL227 IL202
Fig. 2 Maximum Parsimony analysis. PAUP 4.0. Strict consensus of 1000 most parsimonious trees. Description of these 1,000 trees: L = 775, CI = 0.5316, RI = 0.8879. Numbers at nodes refer to bootstrap percentages for 1,000 replicates. Two lineages discussed in text are labeled L1 and L2. Two clades discussed in text are labeled A and B substitutions). On the contrary, populations of D. carpathicus and D. longipes from single restricted study areas provided similar but slightly different haplotypes, as observed for populations from areas RQ, Ecop, ENVL. This suggests that populations’ intermingling occurs, but the extent differs from species to species. The existence of single haplotypes in some restricted study areas suggests a single source. Possibly, starlings from restricted study area IL and swallows from CB have been infested by only one population of D. gallinae and D. hirundinis, respectively, whereas several infestation events may have occurred in areas RQ, Ecop and ENVL, with D. carpathicus and D. longipes. Within the D. gallinae clade, 9 different populations from various geographic origins in France share a single haplotype (layer hen farms: 8009, 8002b, 8003b1, 8006a, F01-5013, F50S, F86, F22AR; wild bird: Percno; cf. Fig. 3). This group is labelled A in Figs. 2 and 3.
Observed host range
Comparison between literature and field data
Table 5 provides an overview of the distribution of Dermanyssus species across bird hosts using published literature data (based on morphological diagnosis of mites) and our field 123 PUBLICATION IV
Exp Appl Acarol (2009) 48:115–142 129
T pyri O bacoti A casalis 1.00 GO54 MAR GO58a D. apodis GO36 D hirsutus D quintus LC23E 0.99 JBO59 1.00 LR20A 1.00 Ecop1 EcopO6-5 Ecop3a 0.98 Ecop06 9a 1.00 RQ D. carpathicus RQ_Mes1 Veol JMC10 LC10A RQ_Mes3 1.00 PAS ENVL08 3 ENVL08 8 ENVL08 1 D. longipes 0.95 HIR6b HIR1 ADhirun 1.00 1.00 0.99 TROAED PARATR TACBIC CB3 CB5 D. hirundinis HIR6a CB4 0.99 OC 0.1 CHOV HR 1.00 LB07 4 JBO517 L2 LB18 0.74 LC 1.00 COL JGC L1 PI CANIM 8008 0.85 F01 5013 Percno 8006a 8003b1 8002b A 8009 F22AR 0.70 F86 F50S ROL1 D. gallinae CANIT JB046 JBO27 8007 Fa2 PO1 PO2 8004 B 8012 8006b Fa1 Chab SB SK Woodp ROL2 IL302 IL213 IL227 IL202
Fig. 3 Bayesian analysis. MrBayes v3.1.2, GTR ? C?i model of evolution for 5 9 106 generations. Numbers at nodes refer to Bayesian Posterior Probabilities. Additional symbols indicate the type of anthropogenic ecosystem: ‘‘w’’ pigeons breeding facilities, ‘‘w’’ canary breeding facilities, ‘‘I’’ layer hen or chicken houses, ‘‘5’’ apple/pear orchards. Populations without any of these symbols have been collected in natura. Two lineages discussed in text are labeled L1 and L2. One clade discussed in text is labeled B. Group A corresponds to clade A in Fig. 2, and groups together populations sharing the same haplotype data (based and on morphological and molecular diagnosis of mites). It includes data on 111 bird species (two of which undetermined: Parus sp. and Passer sp.), distributed over 9 bird orders and 31 bird families. Of these 111 species, data on 69 bird species were derived from the literature only, data on 17 bird species stemmed from literature and our field data and data on 25 bird species originated only from our field samples, which included rather variable numbers of nests per bird species. Twenty-five mite species are currently included in the genus Dermanyssus. Of these 25 species, five have been found in our field samples (D. carpathicus, D. gallinae, D. hir- undinis, D. longipes, D. apodis). D. carpathicus was found in association with four passeriform bird species of two different genera [P. major and P. caeruleus, Phoenicurus phoenicurus (L., 1758) and Ph. ochruros (Gmelin, 1774)]. These host genera were known from the literature. D. gallinae was found in various, distant bird groups. It was previously recorded from 25 bird species, 4 of which were in our field samples [G. gallus, Serinus canaria (L., 1758), P. major, S. vulgaris]. Our field data also provided records of three bird species [Dend- rocopos major (L., 1758), Neophron percnopterus (L., 1758) and C. garrulus] that are new as hosts of D. gallinae. D. hirundinis was found exclusively and frequently (at least 30%) in nests of two species of Hirundinidae, the type host family in France. Since it was previously recorded from almost 40 different bird species distributed in nine different bird orders, we would 123 PUBLICATION IV
130 Exp Appl Acarol (2009) 48:115–142 have expected a wider host range. Recently, some individuals from the restricted study area BMOC (Table 1) were found not only in nests of swallows (Hirundinidae), but also in nests of tits (Paridae) and wrens (Certhiidae) in the USA (cf. ‘‘Host switches via nest sharing of host birds’’). D. longipes has been found in a nest of Passer montanus and in several nests of Parus sp. Its type host was Passer domesticus (L., 1758). Moreover, some mites found by Bra¨nnstro¨m et al. (2008) from several species of Muscicapidae (Passeriforms) provided an ITS sequence similar to that obtained from our populations of D. longipes (Roy et al. 2008).
Host transfer of populations from four Dermanyssus species
To assess the ability of four Dermanyssus species to feed on hosts other than ones they are associated with, bioassays were carried out. These bioassays were not repeated or stan- dardized enough to be dealt with statistically. At best they may give a hint as to the ability to feed and develop on the new host (cf. Tables 2, 3).
D. gallinae A strain of D. gallinae named SK and cultured on hens for the last ten years was, transferred to ducks and canaries, where they readily fed and successfully reproduced. Another population of D. gallinae (8010) also showed such an ability to feed on canaries under laboratory conditions, immediately after having been collected from a layer hen farm (after a starvation period of 4 days). This population shares the haplotype of group A (cf. above), ie provides a haplotype very common in French layer farms.
D. longipes and D. carpathicus D. longipes and D. carpathicus did not reveal an ability to feed on a new host to the extent observed for D. gallinae. D. longipes was unable to develop on a canary in the laboratory (long time bioassay). A population of D. carpathicus caught from a nest of P. ochruros (Gmelin 1774) appeared to maintain itself during about one year on a canary, since brilliant red individuals were regularly noticed. However, short time bioassays with canaries as the new host suggested a very different behaviour in D. carpathicus than in D. gallinae. In the former species, most individuals released were not recovered and there were only a few engorged mites observed, whereas in the latter species, simultaneously under similar conditions, most individuals were found aggregated and engorged.
D. hirundinis Although tested only once in bioassays, D. hirundinis did not seem to be different from D. gallinae in its ability to feed and develop on canaries as a host.
Discussion
Two main clades appear in the species of Dermanyssus tested here. One clade includes D. carpathicus, D. hirundinis, D. longipes, and D. apodis. The second clade includes the various lineages of D. gallinae. Within the D. gallinae clade, two lineages are strongly separated from the others.
D. gallinae: several lineages, some more specialized than others
The two lineages that stand out as strongly isolated from each other and from the other lineages, form a sister group to all other D. gallinae. All analyses provided strong support
123 PUBLICATION IV
Exp Appl Acarol (2009) 48:115–142 131 for the monophyly of populations of L1 (Fig. 3) that were collected mainly from pigeons (domestic pigeons from breeding facilities and a nest sampled in a town) and stem from distant geographic areas. The monophyly of lineage L2 also receives strong support from the analyses shown in Figs. 2 and 4, but is not so strongly supported in other single gene analyses (Roy et al. 2008). Populations in this lineage were only found three times in wild Passeriformes. The sister group of the clade L1 ? L2 groups together populations found in industrial layer hen farms and some other populations (67% bootstrap in MP, 0.74 Bayesian Posterior Probabilities in Bayesian analysis). Within this clade, two major groupings become manifest, that are also found in wild avifauna: one grouping concerns D. gallinae popu- lations in hen farms, in nests of N. percnopterus and on C. garrulus (group A, Figs. 2, 4) and the other grouping concerns D. gallinae populations in hen farms and in nests of S. vulgaris, P. major and D. major and on C. garrulus (clade B, cf. Figs. 2, 4). All popu- lations in group A except ROL1 and CANIT provided the same haplotype, even though these populations are of quite different geographic origin (cf. Appendix). In clade B, approximately one haplotype per population is present, but in some cases there were haplotypes with very small differences: for instance the haplotype of PO1 and PO2 (layer hens from Poland) is very close to that of Fa1 (layer hens from Norway), differing by a single nucleotide substitution.
Assessments on host specificity using observed host range and ecological observations
The clade D. apodis? had a much narrower host range than D. gallinae: only Passeri- formes represent hosts for D. carpathicus, D. longipes and D. hirundinis, Piciformes for the hirsutus-group and Apodiformes in the basal D. apodis, whereas D. gallinae was isolated in eight different bird orders.
Host spectrum enlarging within the most synanthropic clade (D. gallinae)
L1 and L2 appear to be more specific than their sister clade. L1 was mainly found in association with pigeons and L2 with bird species of two different passeriform families. For L1, there were only two records from other bird groups, but these could well be cases of mites that do not actually infest these birds: one dead and dried individual was found in a nest of a predatory bird [Tyto alba (Scopoli 1769)] and another single individual was found dead in a nest of black swifts sampled in the town of Nıˆmes. Since in the town environment pigeons and swifts tend to compete for nesting places, the latter individual may have stemmed from a pigeon host. For these reasons, L1 could be specific to pigeons. L2 was never found in farms or other anthropogenic environments, whereas L1 was found in facilities for pigeon breeding. The remainder of the D. gallinae populations showed very little divergence in mt-Co1 (only 19 haplotypes among 30 populations, pairwise divergence percentage 1–6%, vs. 8– 12% between these populations and L1 or L2), were mainly collected from hens (Galli- formes), but also from completely different bird groups (Coraciiformes, Piciformes, Passeriformes, Ciconiiformes). Moreover, some lab bioassays succeeded in feeding mites of different D. gallinae populations, freshly caught from hen farms, on canaries and on duck chicks and these mites showed a rate of development similar to that on their original host (Tables 2, 3). These D. gallinae populations seem to quickly adapt to new hosts and the large clade to which these populations belong was found exclusively in layer hen farms. 123 PUBLICATION IV
132 Exp Appl Acarol (2009) 48:115–142
Host switches via nest sharing of host birds
The cases where haplotypes were found to be restricted to certain areas (sometimes with several successive samples, year after year; Table 1) require scrutiny because they may help to get insight into a possible mode of dissemination of Dermanyssus mites. For example, 2 sites with D. carpathicus, one with D. longipes, two with D. hirundinis and one with D. gallinae led us to suggest (1) that a single species was present in association with 1 (or 2) bird species in most areas that had a diameter compatible with the home range of birds under consideration, and (2) that mixing of mite populations occurs in restricted study areas in all Dermanyssus species under test. Thus, often a single haplotype was found per restricted study area, or haplotypes differing by only a few nucleotide substitutions. In 2 nests, however, two different populations were noted in the same Dermanyssus species (HIR6a and HIR6b both belonging to D. hirundinis, RQ-mes1 and RQ-mes3 both belonging to D. carpathicus). The concerned bird species (tits in genus Parus, starlings, swallows) are known to spend all time in a restricted area throughout the year and to rebuild their nest upon an older one, either their own or that of another (Caparros, Bouvier, Personal communication). Thus, there is ample opportunity for the Dermanyssus mites to switch from one host to another in case their hosts share nesting places, as suggested by Valera et al. (2003). Contact between nest and bird seems to be absolutely necessary. This mode of dis- semination via nest sharing is supported by our results in that different host ranges were noted in D. hirundinis between France (populations HIR6a, HIR6b, HIR1, CB3, CB4, CB5, OC, CHOV, HR) and the USA (populations ADhirun, TROAED, PARATR, TAC- BIC) (cf. Figs. 2, 3, and appendix for host affiliation). Indeed, this species has been isolated exclusively from H. rustica and D. urbicum (Hirundinidae) in France, whereas it was found from three different passeriform families in the USA (Hirundinidae, Certhiidae, Paridae). But the American hirundinid species tested is not present in France (Tachycineta bicolor (Vieillot 1808)) and its ecology strongly differs from the 2 tested French species: T. bicolor or the Tree Swallow is a cavity nesting bird and often uses nestboxes in the USA, as do the two other American host species (Poecile atricapillus L., 1766, the Black-Capped Chickadee and Troglodytes aedon Vieillot, 1809, the House Wren). These three bird species are often found sharing the same nestboxes in the USA (O. Dehorter, Personal communication), whereas none of the two tested French hirundinid species are found sharing nestboxes in France. This opportunistic behaviour of D. hirundinis is also observed in our bioassays. Indi- viduals of D. hirundinis directly sampled from a fresh nest of H. rustica did not show differences with individuals of D. gallinae in engorgement and development on canaries as hosts. The apparent host specificity observed in France is therefore likely the result of ecological and/or geographic factors. Alternatively, there may be genetic differences between the French and American populations tested, as indicated by small differences in COI sequences. Evidence strongly suggests that dissemination of mites happens in cases of nest sharing between congeners and different species. The ability to switch from one host to another at a distance of several meters seems to be greatly reduced at least in some of these species. As indicated by scrutiny of the population in the restricted area CB (Table 1), some D. gallinae may rather starve for several weeks to a few months, than to venture bridging the distance to new hosts only metres away. For example, in absence of chicken, D. gallinae individuals present in large numbers in chicken cages do not appear to move to nests of swallows nearby. In these swallow nests, only D. hirundinis was found! 123 PUBLICATION IV
Exp Appl Acarol (2009) 48:115–142 133
The nest seems to be a reservoir and the bird host could be the carrier. Note that all Dermanyssus individuals directly collected on flying birds in the present study (cf. Appendix) were adult females. Is this a stage adapted to dissemination by phoresy? Note also that, Flechtmann and Baggio (1993) reported one case of D. gallinae phoretic on a beetle, and the isolated individuals were adult females.
Synanthropicity and nest microenvironment
The micro-environmental conditions of farms and breeding facilities are likely to differ from those in wild bird nests. For example, these anthropogenic environments harbour large numbers of bird individuals in a small area, provide relatively regular temperatures and humidity, are usually exposed to some pesticides and provide hosts during most of the year (layer hen houses are usually bird-free for less than 2 months per year). On the contrary, in wild bird environments, even in cases of bird colonies, the number of bird individuals and the area they occupy is much smaller than in farms, and temperature and humidity are much more variable during the year and even during the day, than in farms. Also, pesticide products are typically absent (except in the case of orchards in restricted study area JBO) and the host is available only during a limited period, i.e. the breeding period of their host in spring-summer, and sometimes winter nights (implying absence of bird host for several months in autumn and summer). The two main clades in the Dermanyssus species tested exhibit conspicuous adaptations to their microenvironments. Clade D. apodis? was never found in any farm or breeding facility, but clade D. gallinae was present in poultry farms (layer hens, chickens) or breeding facilities for canaries, other Fringillidae and pigeons, where they usually prolif- erate. Moreover, the latter clade was found in agroecosystems such as orchards. Taken together, it can be concluded that the clade D. gallinae is unique in harbouring synan- thropic populations (Fig. 3).
Role of the ecosystem in the nest environment
The nest provides a specific environment shaped by various organisms together forming a micro-ecosystem. First and foremost are the birds occupying the nest. They bring nest- building material and release waste products, but they may also remove them. So nest building and nest hygiene may be determinants of the nest as a biotope for Dermanyssus spp. In wild birds, D. gallinae was found proliferating only in pigeon nests (L1) and starling nests (restricted study area IL), which are birds that allow droppings from chicks to dry within the nest. Moreover, D. gallinae is commonly found in layer hen farms where hen droppings accumulate around the flock and numerous D. gallinae individuals are often found aggregating under dried droppings. In contrast, D. hirsutus, D. quintus, D. carpa- thicus, D. longipes and D. hirundinis, in clade D. apodis ?, were never found in such nests and seem to proliferate only in passeriform or piciform nests which are regularly cleaned by the parental birds. One may wonder how the presence of bird droppings affects species of clade D. apodis ? relative to D. gallinae. Another factor affecting the proliferation of Dermanyssus spp. could be the presence or absence of pesticide products. Results from nests of restricted study area JBO suggest that D. gallinae is better adapted to an orchard agroecosystem than D. carpathicus: the latter was found in natura (either in natural nest or in nestbox) in 15% of the Parus
123 PUBLICATION IV
134 Exp Appl Acarol (2009) 48:115–142 nests sampled in France, whereas only once in fruit orchards (62 nests; i.e. 1,6%). However, D. gallinae was found in tit nests only when the nests were located in orchards (Fig. 1). Finally, various other organisms inhabit nests, these are mostly insects and mites and represent different feeding guilds: bird parasites, predators, microbivores and scavengers. It is interesting to observe that tree-nesting birds generally suffer from attack by Derma- nyssus spp. and that the breeding success of individual birds strongly depends on the presence of predatory mites that attack Dermanyssus spp. (Lesna et al. 2009). The only ground nesting group extensively analysed here are the larks (Alaudidae) and these bird species appear to stand out as the only ones that had no Dermanyssus individuals in their nests (Fig. 1). As a rule, these nests contained many species of predatory arthropods (data not shown) and the presence of various predators was inversely proportional to the pres- ence of D. gallinae. The case of restricted study area MOL is a good illustration of it, with only one D. gallinae individual found among several cubic meters of litter sampled, but several species of predators present in considerable numbers.
Perspectives
To understand the distribution of Dermanyssus spp. and lineages over bird species as hosts, co-phylogenetic analysis is of fundamental importance (especially with respect to D. carpathicus, D. hirundinis, D. longipes, D. apodis and their bird hosts). However, the presence of a Dermanyssus species in the nest of a bird species depends on several ecological factors that may dramatically alter the potential host range. These include bird- related factors, such as nest material selection, nest hygiene, competition for nest sites and nest sharing, but also the presence of various guilds of nidicolous arthropods, including predators of Dermanyssus spp. and flying insects that may act as vectors for dissemination of Dermanyssus spp., being wingless and therefore less mobile. Given that anthropogenic environments, such as poultry farms, offer an environment that is widely different from that of wild birds, it is to be expected that Dermanyssus spp. (especially D. gallinae) undergo strong selection to adapt to this environment (e.g. by developing pesticide resistance). It is the extent to which these adaptations are decisive for survival in the anthropogenic environment and the extent to which they affect survival in the natural environment, which will determine how much exchange there will be between Derma- nyssus from wild and domestic birds. Elucidating these adaptations will therefore be an important task for future research.
Acknowledgments Authors would like to warmly thank for help, sampling, advices: O. Bain (Muse´um National d’Histoire naturelle (MNHN), Paris, France), Y. Beauvallet (Centre de Recherche sur la Biologie et les Populations d’Oiseaux (CRBPO), MNHN, Paris, France), S. Bonnet (LEGTA, Saint-Genis Laval, France), J. C. Bouvier (INRA unite PSH, Avignon, France), L. Brucy (CRBPO, MNHN, Paris, France), O. Caparros (CRBPO, MNHN, Paris, France), L. Carrier (CRBPO, MNHN, Paris, France), F. Humbert (CRBPO, MNHN, Paris, France), T. Cencek, (Pan´stwowy Instytut Weterynaryjny, Puławy, Poland) M. De Rojas (Universidad de Sevilla, Sevilla, Spain), P. Chabrol (veterinary office, Bourg-en-Bresse, France), J. M. Chavatte (MNHN, Paris, France), O. Dehorter (CRBPO, MNHN, Paris, France), J. Delaporte (Bayer Animal HealthCare, Saint-Ave, France), A. G. Gjevre (National Veterinary Institute of Norway, Oslo, Norway), G. Gory (Museum d’Histoire naturelle, Nıˆmes, France), N. Guichard (LEGTA, Saint-Genis Laval, France), P. A. Heuch (National Veterinary Institute of Norway, Oslo, Norway), G. Inizan (Bayer Animal HealthCare, Saint-Ave, France), O. Kilpinen (Danish Institute of Agricultural Sciences, Lyngby, Denmark), S. Kreiter (Supagro, Montpellier, France), G. Lallemand (Lyce´e des Mandailles, Chaˆteauneuf de Galaure, France), B. M. OConnor (Museum of Zoology, University of Michigan, Ann Arbor, USA), B. Paoletti
123 PUBLICATION IV
Exp Appl Acarol (2009) 48:115–142 135
(University of Teramo, Teramo, Italy), M. Rigaux (IUT A, Universite´ Lyon1, Villeurbanne, France), P. Tavernier (Centre de Soins aux Oiseaux sauvages du Lyonnais, Francheville, France), F. Veau (CORA Arde`che, France), Vincent-Martin N. (CRBPO, MNHN, Paris, France), J. Witters (Institute for Agricultural and Fisheries Research, Merelbeke, Belgium).
Glossary Mono-/polyxeny Condition of host specificity for a parasite species that needs a single host species/several host species for completion of its development. Synanthropic Ecologically associated with humans.
Appendix
See Fig. 4 and Table 6.
Fig. 4 Location of French departments. In grey are highlighted French departments in which Dermanyssus individuals have been found during field sampling in the present study (2005–2007)
123 3 x plAao 20)48:115–142 (2009) Acarol Appl Exp 136 123 Table 6 List of DNA-tested mite populations, including accession numbers
Pop code Species of mite 18S-28S 16S rRNA COI Country Restricted Sampling year Host Context rRNA (State, French study area (restricted ‘department’) study areas)
8002 D. gallinae FM208713 France, 26 Gallus gallus (Galliform) Layer farm 8003 D. gallinae FM208733 France, 38 Gallus gallus (Galliform) Layer farm 8004 D. gallinae FM208722 France, 69 Gallus gallus (Galliform) Little amateur hen house 8005 D. gallinae FM208737 Gallus gallus (Galliform) Little layer and chicken farm 8006 D. gallinae FM208725 and France, 01 Gallus gallus (Galliform) Layer farm FM208732 8007 D. gallinae FM208717 Belgium Gallus gallus (Galliform) Layer farm 8008 D. gallinae FM208712 France, 69 Columba livia Breeding facilities for (Columbiform) bird competitions 8009 D. gallinae FM208724 France, 69 Gallus gallus (Galliform) Layer farm 8012 D. gallinae FM208739 France, 01 CB 2008 (summer) Gallus gallus (Galliform) Cages with chickens 8010 D. gallinae FM881897 France, 26 Gallus gallus (Galliform) Layer farm ACA A. casalis AM903317 AM921907 AM921868 France, 69 - Breeding facility ADhirs D. hirsutus AM931077 AM921912 AM921878 USA, MI Colaptes cafer (Piciform) On bird ADhirun D. hirundinis AM931076 AM921913 AM921881 USA, MI BMOC ? Tachycineta bicolor On bird (Passeriform) ADqui D. quintus AM931075 AM921882 USA, MI Picoides villosus (Piciform) On bird CANIM D. gallinae FM208734 France, 69 (Pet shop with various bird Just caught from a cage species very close to each with canaries and other) quails; many other bird species in cages next to them (including Psittaciforms, …) PUBLICATION IV CANIT D. gallinae AM903308 AM921909 AM921877 Italy Serinus canarius Breeding facility (Fringillidae: Passeriform) x plAao 20)4:1–4 137 48:115–142 (2009) Acarol Appl Exp Table 6 continued
Pop code Species of mite 18S-28S 16S rRNA COI Country Restricted Sampling year Host Context rRNA (State, French study area (restricted ‘department’) study areas)
CB3 D. hirundinis FM208726 France, 01 CB 2008 (summer) Hirundo rustica Nest in farm building (Passeriform) CB4 D. hirundinis FM208727 France, 01 CB 2008 (summer) Hirundo rustica Nest in farm building (Passeriform) CB5 D. hirundinis FM208728 France, 01 CB 2008 (summer) Hirundo rustica Nest in farm building (Passeriform) Chab D. gallinae AM931074 AM921886 AM921857 France, 01 Gallus gallus (Galliform) Layer farm CHOV D. hirundinis AM943019 FM179369 France, 72 Hirundo rustica Nest in a barn (Passeriform) COL* D. gallinae AM903307 AM921892 AM921875 France, 69 Columbus livia On adult bird (Columbiform) DR D. gallinae AM931073 AM921885 Spain Fringillidae (Passeriform) Cage Ecop1 D. carpathicus FM208731 France, 42 Ecop 2008 (winter) Parus sp. (Passeriform) Nest box Ecop3 D. carpathicus FM208729 France, 42 Ecop 2008 (winter) Parus sp. (Passeriform) Nest box Ecop06-5 D. carpathicus AM903314 AM921901 AM921873 France, 42 Ecop 2006 (summer) Parus major (Passeriform) Nest box Ecop06-9 D. carpathicus FM208730 France, 42 Ecop 2007 (summer) Parus caeruleus Nest box ENVL07-07 D. longipes France, 69 ENVL 2007 (summer) Parus sp. (Passeriform) Nest box ENVL08-1 D. longipes FM208744 France, 69 ENVL 2008 (summer) ENVL08-3 D. longipes FM179377 FM179374 FM179365 France, 69 ENVL 2008 (summer) Parus sp. (Passeriform) Nest box ENVL08-7 D. longipes ENVL 2008 (summer) Parus sp. (Passeriform) Nest box ENVL08-8 D. longipes FM208743 France, 69 ENVL 2008 (summer) Parus sp. (Passeriform) Nest box F01-5013 D. gallinae FM208721 France, 01 Gallus gallus (Galliform) Layer farm (free range) F22AR D. gallinae FM208720 France, 22 Gallus gallus (Galliform) Layer farm PUBLICATION IV 123 F50S D. gallinae FM208719 France, 50 Gallus gallus (Galliform) Layer farm F86 D. gallinae FM208718 France, 86 Gallus gallus (Galliform) Layer farm Fa1 D. gallinae AM931072 AM921884 AM921853 Norway Gallus gallus (Galliform) Layer farm Fa2 D. gallinae AM931071 AM921883 AM921852 Norway Gallus gallus (Galliform) Layer farm (organic) 3 x plAao 20)48:115–142 (2009) Acarol Appl Exp 138 123 Table 6 continued
Pop code Species of mite 18S-28S 16S rRNA COI Country Restricted Sampling year Host Context rRNA (State, French study area (restricted ‘department’) study areas)
GO1 D. apodis AM903299 AM921894 France, 30 Apus apus (Apodiform) Nest GO10 D. apodis AM921895 France, 30 Apus apus (Apodiform) Nest GO12 D. apodis AM903309 France, 30 Apus apus (Apodiform) Nest GO15 D. apodis AM903313 AM921896 France, 30 Apus apus (Apodiform) On young birds at nest GO16 D. apodis AM903313 France, 30 Apus apus (Apodiform) On young birds at nest GO26 D. apodis AM921900 France, 30 Apus apus (Apodiform) Nest GO36 D. apodis FM179371 France, 30 Apus apus (Apodiform) Nest GO44 D. apodis AM921898 France, 30 Apus apus (Apodiform) Nest GO46 D. apodis AM921897 France, 30 Apus apus (Apodiform) Nest GO54 D. apodis AM930888 AM921874 France, 30 Apus apus (Apodiform) Nest GO58a D. apodis FM179370 France, 30 Apus apus (Apodiform) Nest GO8* D. gallinae AM921893 France, 30 Apus apus (Apodiform) Nest HIR1 D. hirundinis FM179379 FM179366 France, 85 HIR 2008 (winter) Hirundo rustica Nest (Passeriform) HIR2 D. hirundinis France, 85 HIR 2008 (winter) Hirundo rustica Nest (Passeriform) HIR5 D. hirundinis France, 85 HIR 2008 (spring) Hirundo rustica Nest (Passeriform) HIR6 D. hirundinis FM208741 and France, 85 HIR 2008 (spring) Hirundo rustica Nest FM208740 (Passeriform) HR D. hirundinis AM903300 AM921888 AM921872 France, 69 Hirundo rustica Nest in farm building (Passeriform) IL302 D. gallinae FM207495 The Netherlands IL 2007 Sturnus vulgaris Nest box
(Passeriform) PUBLICATION IV IL213 D. gallinae FM207490 FM207492 FM207499 The Netherlands IL 2007 Sturnus vulgaris Nest box (Passeriform) IL227 D. gallinae FM207491 FM207494 FM207496 The Netherlands IL 2007 Sturnus vulgaris Nest box (Passeriform) x plAao 20)4:1–4 139 48:115–142 (2009) Acarol Appl Exp Table 6 continued
Pop code Species of mite 18S-28S 16S rRNA COI Country Restricted Sampling year Host Context rRNA (State, French study area (restricted ‘department’) study areas)
IL202 D. gallinae FM207497 and The Netherlands IL 2007 Sturnus vulgaris Nest box FM207498 (Passeriform) JBO27 D. gallinae FM208716 France, 13 Parus major (Passeriform) Nest box in an apple orchard JBO46 D. gallinae FM208736 France, 13 Parus major (Passeriform) Nest box in an apple orchard JBO517 D. gallinae AM930885 AM921879 France, 13 Parus major (Passeriform) Nest box in an apple orchard JBO59 D. carpathicus AM930882 AM921902 AM921870 France, 13 Parus major (Passeriform) Nest box in an apple orchard JMC10 D. carpathicus AM943018 AM943021 France, 62 Parus major (Passeriform) Nest LB074 D. gallinae AM921866 France, 18 LB 2006-2007 Delichon urbica Nest (Passeriform) LB18 D. gallinae AM930889 AM921908 AM921867 France, 18 LB 2006-2007 Delichon urbica Nest (Passeriform) LC D. gallinae AM903306 AM921891 AM921859 France, 26 Columbus livia Breeding facility, rural (Columbiform) country LC23 D. carpathicus FM208735 and France, 26 Parus major (Passeriform) Nest FM881898 LC10A D. carpathicus FM179367 Parus major (Passeriform) Nest LR20A D. carpathicus FM179368 Parus sp. (Passeriform) Nest MAR D. apodis AM945880 AM921899 AM921880 France, 69 Apus apus (Apodiform) On young bird fallen from nest Ob O. bacoti AM903318 AM921905 FM179677 ? rodents From a lab strain in MNHN (O. Bain, Lab of Parasitology) P 123 OC D. hirundinis AM903312 AM921889 AM921862 France, 38 Delichon urbica On young birds at nest UBLICATION I (Passeriform) V 4 x plAao 20)48:115–142 (2009) Acarol Appl Exp 140 123 Table 6 continued
Pop code Species of mite 18S-28S 16S rRNA COI Country Restricted Sampling year Host Context rRNA (State, French study area (restricted ‘department’) study areas)
PARATR D. hirundinis FM208746 USA, MI BMOC 2007 (autumn) Parus atricapillus Nest box ? (Passeriform) Parm D. carpathicus AM903315 France, 69 Parus major (Passeriform) Nest box PAS D. longipes AM903310 AM921904 AM921869 France, 13 Passer montanus Nest (Passeriform) Percnobis D. gallinae AM943020 FM208738 France, 07 Neophron percnopterus Nest (Ciconiiform) PI* D. gallinae FM179378 FM179375 AM921860 France, 13 Columbus livia From nest inside a flat in (Columbiform) city center PO1 D. gallinae AM903302 AM921854 Poland Gallus gallus (Galliform) Layer farm PO2 D. gallinae AM921914 AM921855 Poland Gallus gallus (Galliform) Layer farm ROL1 D. gallinae AM903304 AM921910 AM921864 France, 13 Coracias garrulus On adult birds (Coraciiform) ROL2 D. gallinae AM903305 AM921911 AM921865 France, 13 Coracias garrulus On young birds at nest (Coraciiform) RQ D. carpathicus AM903316 AM921903 AM921876 France, 42 RQ 2006 (summer) Phoenicurus phoenicurus Nest, near a house at (Passeriform) altitude of ca. 1000 m RQ-Mes D. carpathicus FM208715, France, 42 RQ 2006 (summer) Parus sp. (Passeriform) Nest, in the wall of a FM208714 house, alt ca. 1000 m and FM208723 SB D. gallinae AM921858 France, 69 Gallus gallus (Galliform) Small, amateur hen house SK D. gallinae AM903303 AM921887 AM921856 Denmark Gallus gallus (Galliform) Layer farm TACBIC D. hirundinis FM208745 USA, MI BMOC 2007 (autumn) Tachycineta bicolor Nest box (Passeriform) PUBLICATION IV TROAED D. hirundinis FM208747 USA, MI BMOC 2007 (autumn) Troglodytes aedon Nest box (Passeriform) x plAao 20)4:1–4 141 48:115–142 (2009) Acarol Appl Exp Table 6 continued
Pop code Species of mite 18S-28S 16S rRNA COI Country Restricted Sampling year Host Context rRNA (State, French study area (restricted ‘department’) study areas)
TPYR T. pyri FM179376 FM179373 FM179364 France – From a lab strain in Supagro, Montpellier (S. Kreiter) Veol D. carpathicus AM921871 France, 69 Parus major (Passeriform) Nest box Woodp D. gallinae AM903301 AM921890 AM921863 France, 69 Dendrocopos major On wild adult female (Piciform) bird
Each population corresponds to a group of Dermanyssus mites from a single nest (or a single building in a farm or breeding facility). There was never more than 1 species of Dermanyssus per nest P 123 UBLICATION I V PUBLICATION IV
142 Exp Appl Acarol (2009) 48:115–142
References
Bra¨nnstro¨m S, Morrison DA, Mattsson JG, Chirico J (2008) Genetic differences in internal transcribed spacer 1 between Dermanyssus gallinae from wild birds and domestic chickens. Med Vet Entomol 22(2):152–155. doi:10.1111/j.1365-2915.2008.00722.x Carroll H, Beckstead W, OConnor T, Ebbert M, Clement M, Snell Q, McClellan D (2007) DNA reference alignment benchmarks based on tertiary structure of encoded proteins. Bioinformatics 23(19):2648– 2649. doi:10.1093/bioinformatics/btm389 De Lillo E (2001) A modified method for Eriophyoid mite extraction (Acari: Eriophyoidea). Int J Acarol 27(1):67–70 De Souza Amorim D (1982) Classificac¸ao por sequenciac¸ao: uma proposa para a denominac¸ao dos ramos retardados. Rev Bras Zool 11:1–9 Flechtmann HW, Baggio D (1993) On Phoresy of hematophagous ectoparasitic acari (Parasitiformes: Ixodidae and Dermanyssidae) on Coleoptera observed in Brazil. Int J Acarol 19(2):195–196 Hoogstraal H, Aeschlimann A (1982) Tick host specificity. Mitteilungen des Schweizerischen entomolog- ischen Gesellschaft 55:5–32 Huelsenbeck JP, Ronquist F (2001) MRBAYES: Bayesian inference of phylogeny. Bioinformatics 17: 754–755. doi:10.1093/bioinformatics/17.8.754 Knee W (2008) Five new species of Rhinonyssidae (Mesostigmata) and one new species of Dermanyssus (Mesostigmata: Dermanyssidae) from birds of Alberta and Manitoba, Canada. J Parasitol 94:348–374. doi:10.1645/GE-1358.1 Lesna I, Wolfs P, Faraji F, Roy L, Komdeur J, Maurice W. Sabelis (2009) Candidate predators for biological control of the poultry red mite Dermanyssus gallinae. Exp Appl Acarol. doi:10.1007/s10493-009- 9239-1 Livezey BC, Zusi RL (2007) Higher order phylogeny of modern birds (Theropoda, Aves: Neornithes) based on comparative anatomy II. Analysis and discussion. Zool J Linn Soc 149:1–95. doi:10.1111/ j.1096-3642.2006.00293.x Moss WW (1978) The mite genus Dermanyssus: a survey, with description of Dermanyssus trochilinis,n. sp., and a revised key to the species (Acari: Mesostigmata: Dermanyssidae). J Med Ent 14(6):627–640 Nylander JAA (2004) MrModeltest v2. Program distributed by the author. Evolutionary Biology Centre, Uppsala University Peterson AP (2008) http://www.zoonomen.net/avtax/frame.html, Version 7.07 (2008.07.30) Ronquist F, Huelsenbeck JP (2003) MRBAYES 3: Bayesian phylogenetic inference under mixed models. Bioinformatics 19:1572–1574. doi:10.1093/bioinformatics/btg180 Roy L, Chauve CM (2007) Historical review of the genus Dermanyssus Duge`s, 1834 (Acari: Mesostigmata: Dermanyssidae). Parasite 14(2):87–100 Roy L, Dowling APG, Chauve CM, Buronfosse T (2008) Delimiting species boundaries within Derma- nyssus Duge`s, 1834 (Acari: Mesostigmata) using a total evidence approach. Mol Phylogenet Evol. doi: 10.1016/j.ympev.2008.11.012 Swofford DL (2001) PAUP*–Phylogenetic Analysis Using Parsimony (*and other methods). Version 4.0B 10. [Computer software and manual]. Sinauer Associates, Sunderland, Massachusetts Valera F, Casas-Criville´ A, Hoi H (2003) Interspecific Parasite exchange in a mixed colony of birds. J Parasitol 89(2):245–250 Valiente Moro C, Chauve C, Zenner L (2005) Vectorial role of some dermanyssoid mites (Acari, Meso- stigmata, Dermanyssoidea). Parasite 12(2):99–109 Valiente Moro C, Fravalo P, Amelot M, Chauve C, Zenner L, Salvat G (2007) Colonization and invasion in chicks experimentally infected with Dermanyssus gallinae contaminated by Salmonella enteritidis. Avian Pathol 36(4):307–311. doi:10.1080/03079450701460484 Wilm A, Mainz I, Steger G (2006) An enhanced RNA alignment benchmark for sequence alignment programs. Algorithms Mol Biol 1(19):1–11
123 5.2 Diversité génétique et flux de populations chez quelques espèces du groupe gallinae : publication V
a - Présentation De concert avec divers paramètres écologiques, des caractéristiques intrinsèques président à l’adaptabilité des organismes à de nouveaux environnements. Ainsi, dans un taxon donné, tant animal que végétal ou fongique, certaines entités peuvent posséder une plus grande capacité à coloniser un habitat différent de l’habitat (ou des habitats) usuel(s) que d’autres. C’est ainsi que, sous l’effet de modifications induites par l’homme (transformation des écosystèmes, échanges à de grandes échelles géographiques, …), certains organismes révèlent un potentiel invasif important, alors que des entités apparentées demeurent confinées dans leur environnement originel. Des organismes tels l’algue Caulerpa taxifolia ou la grenouille-taureau Rana catesbeiana importées volontairement ou par accident envahissent de ce fait brutalement des aires géographiques qui en étaient exemptes, aux dépens des espèces autochtones des écosystèmes concernés, alors que d’autres organismes importés n’ont pas réussi à s’adapter au nouvel environnement. Selon Williamson (1996), qui distingue quatre étapes dans le phénomène invasif (importation, introduction, établissement, développement nuisible, ie invasion), env. 0,1% des espèces importées deviennent invasives. Les systèmes hôte-parasite, modèles de prédilection pour l’étude des phénomènes adaptatifs, offrent de nombreux exemples de diversité dans les aptitudes adaptatives. En effet, là encore, certaines entités parasites étroitement apparentées diffèrent par le niveau de leur spécificité d’hôte, conséquence de caractéristiques écologiques et comportementales de l’hôte, mais aussi de particularités intrinsèques du parasite.
La biologie microprédatrice et nidicole des espèces du groupe gallinae en font des parasites particulièrement dépendants de l’écosystème du nid et a priori faiblement liés à l’hôte lui-même. Pourtant, une spécificité plus élevée que prévu a pu être notée chez les espèces rustiques dans la publication IV. Cela contraste en outre fortement avec le caractère généraliste de D. gallinae. Certains traits de l’écologie des oiseaux hôtes, influant sur les possibilités de transfert d’une espèce d’hôte à l’autre et sur la composition du microécosystème du nid, sont bien sûr en lien avec ces différences de spécificité. Mais cela ne semble pas expliquer complètement l’ampleur des différences dans les spectres d’hôte observés et le spectre des types d’habitats colonisés (en particulier, sauvages vs anthropisés).
5.2.a.1 Objectifs Les objectifs de la publication V se répartissaient en 2 groupes : (1) Evaluer l’utilité d’un marqueur nucléaire nouvellement développé (Tropomyosin exon n, intron n and exon n+1) pour l’exploration phylogénétique des relations inter- et intraspécifiques dans le groupe gallinae et comparer les résultats obtenus à ceux de la publication III. (2) Répondre aux questions suivantes au niveau spécifique : a) Les lignées généralistes de D. gallinae représentent-elles en fait des espèces cryptiques (potentiellement aussi spécifiques que les espèces « rustiques ») ? b) Est-ce que les généralistes possèdent une plus grande flexibilité évolutive que les spécialistes dans le groupe gallinae ?
125 c) Quel sont les rôles respectifs des flux commerciaux et des échanges entre oiseaux sauvages et volaille domestique dans la dissémination des populations de D. gallinae?
5.2.a.2 Matériel et méthodes Une combinaison d’analyses généalogiques basées sur des haplotypes (comparaison des arbres de gènes, méthode ABC d’investigation de l’histoire démographique des populations basée sur un modèle de type IM, Isolation with Migration), de tests basés sur la théorie de la coalescence et de tests statistiques du polymorphisme génotypique a été mise en œuvre. Des portions de deux loci indépendants – la mt-Co1 (mitochondriale) et la Tropomyosine (nucléaire, portion majoritairement intronique) – d’isolats d’origines géographiques et écologiques différentes ont constitué les matrices de travail. Pour les analyses phylogénétiques, les haplotypes de tous les isolats séquencés ont été inclus quel que soit le nombre d’individus séquencés par isolat. Pour les autres analyses, seuls des isolats dont un nombre relativement important d’individus ont été séquencés (18-24 individus par isolat) ont été pris en compte afin d’être à même d’évaluer la diversité intra-isolat, nécessaire à l’estimation de la diversité interpopulation. Ainsi un isolat de D. apodis et cinq de D. gallinae (un sauvage, un maintenu en laboratoire depuis 12 ans, trois prélevés dans des élevages de pondeuses et un dans un élevage de pigeons) ont pu être intégrés aux tests statistiques. En outre, un « pseudo- isolat » constitué d’une vingtaine d’individus provenant de nids d’hirondelles rustiques collectés en plusieurs endroits en France, appartenant à D. hirundinis a apporté un complément d’information statistique. Une mise en relation récurrente des résultats obtenus avec des traits d’histoire de vie déjà connus ou nouvellement observés a participé à l’interprétation des données obtenues par les différentes analyses. L’intégration d’isolats de D. gallinae ainsi que de quatre autres espèces françaises a permis la comparaison de variations intraspécifiques entre espèces proches tant sur le plan phylogénétique qu’écologique.
5.2.a.3 Principaux résultats Les séquences isolées des deux gènes (mt-Co1 et Tropomyosine) ont montré des polymorphismes comparables entre eux en fonction des espèces. La diversité haplotypique entre isolats de D. gallinae et isolats de D. apodis et D. hirundinis prélevés in natura s’est avérée significativement plus élevée chez la première espèce dans les deux gènes. Le nombre d’haplotypes et la diversité haplotypique au sein des isolats correspondaient bien aux valeurs attendues suivant les lois de la coalescence. Les arbres de gènes mitochondriaux et nucléaires ont montré une incongruence quant au point d’enracinement par les outgroups, mais les groupements spécifiques sont identiques et l’ordre des branchements entre les différentes espèces comparable. Une résolution beaucoup plus importante des relations interspécifiques qu’avec les ITS (Publication III) a été notée avec la Tropomyosine. Les événements de recombinaison semblent relativement peu nombreux sur la portion analysée. Plusieurs sites à insertion/délétion jalonnent la séquence. L’information phylogénétique qu’ils apportent a semblé indéniable (congruence des topologies obtenues en maximum de parcimonie avec gaps exclus (CI=0,80 RI=0,93) et avec une matrice codant les insertions/délétions seules (CI=0,72 RI=0,94)). Ils sont en outre marqués par une variation intraspécifique. Une polyphylie est apparue dans l’arbre des séquences de Tropomyosine chez D. longipes (2 lignées indépendantes). Des informations contenues par les séquences d’ITS obtenues dans le cadre
126 de la publication III corrélées à deux séquences publiées en 2008 par Brännstörm et al. ont confirmé la présence d’une espèce cryptique. L’inclusion de D. hirsutus dans l’analyse a permis d'éliminer une irrésolution notée dans la publication III quant à la position de cette espèce (séparée, quoique sans support, de D. quintus sur la base des ITS, nucléaires, et non des marqueurs mitochondriaux). Le polymorphisme très faible des ITS explique sans doute cette irrésolution. La reconstruction basée sur la Tropomyosine a confirmé sa position proche de D. apodis, avec des valeurs de supports significatives. Cela est venu renforcer l’invalidation de la division groupe gallinae / groupe hirsutus de Moss évoquée dans la publication III. Le groupe hirsutus s’embranche manifestement, aussi bien sur la base des séquences mitochondriales que nucléaires, au beau milieu du groupe gallinae. La Tropomyosine permet une résolution plus importante des relations interspécifiques que l’ensemble des autres gènes testés avec un taux d’homoplasies réduit (CI=0,8238, RI=0,9320 avec les gaps considérés comme 5ème état ; CI=0,8023, RI=0,9263 gaps exclus). La topologie retenue (Fig. 7), suivant le critère du maximum de parcimonie (gaps considérés comme un 5ème état), dessine une organisation proche de celle suggérée par la mt-Co1, mais le clade des espèces « rustiques » (clade A mitochondrial) se révèle paraphylétique. Toutefois, aucune contradiction dans l’ordre de branchement des entités spécifiques n’a été détectée. Le clade (D. gallinae + D. apodis) constitue l’entité distale, la plus dérivée.
127 D. longipes EN D. longipes PAS D. hirundinis
D. carpathicus
D. apodis D D. gallinae s. str. part
D. gallinae L1
D. gallinae s. str. part
Figure 7. Topologie finale retenue (consensus strict). Tropomyosine, maximum de parcimonie (PAUP 4.0), gaps traités comme un 5ème état. Consensus strict de 264 arbres équiparcimonieux, L=1288, CI=0,8238, RI=0,9320.Les numéros aux nœuds représentent, dans l’ordre, les valeurs de bootstrap obtenues après 1000 réplicats / les indices de Bremer relatifs.
Par ailleurs, les données de polymorphisme intra-isolat, de différenciation inter-isolat et les topologies phylogénétiques ont convergé vers la mise en évidence d’une radiation suivie d’hybridations qui semble avoir eu lieu peu de temps après la spéciation entre D. gallinae et D. apodis, dans la lignée D. gallinae. Dans la faune sauvage, le polymorphisme intra-isolat très élevé et la différenciation relativement faible entre isolats sur la base des séquences théoriquement moins dérivées de Tropomyosine (nucléaire) contrastent avec le polymorphisme intra-isolat réduit et la forte différenciation entre isolats de la mt-Co1 (mitochondriale). Les analyses utilisant la méthode ABC ont aussi permis de mettre en évidence des flux de gènes anciens plus importants que les contemporains. Les différences théoriques du taux de mutation de ces deux loci, corrélées à la taille
128 efficace de la population Ne des génomes mitochondriaux et nucléaires, ont permis d’interpréter l’image générée par la mt-Co1 comme plus récente que celle résultant de l’analyse des séquences de Tropomyosine. L’homogénéité relative des séquences de mt-Co1 témoigne d’une différenciation récente des isolats les uns par rapport aux autres. L’hétérogénéité importante des séquences de Tropomyosine témoigne d’un brassage ancien de populations longtemps isolées entre elles. Cela a suggéré une hybridation entre plusieurs espèces naissantes, antérieure à l’isolement signalé par la mt- Co1. La persistence plus ou moins marquée de deux lignées (L1, L3) au sein de ce complexe entre les deux loci, la multiplicité des lignées mitochondriales contenant des séquences de Tropomyosine très divergentes et la mise en évidence par la méthode ABC de flux de gène ancien plus important que les modernes soutiennent l’hypothèse d’une radiation ancienne, avortée ultérieurement par des hybridations multiples. Les hybridations postérieures à la radiation sont très probablement la conséquence de l’homogénéisation des milieux par l’action humaine. D’une manière générale, en brisant des frontières écologiques, celle-ci peut aboutir à la mise en contact de lignées récemment séparées, dont l’isolement reproducteur est encore réversible (Seehausen et al. 2007). Des événements d’hybridation non seulement interrompent le processus de spéciation en cours, mais encore, en produisant de nouvelles combinaisons de gènes, qui plus est dans un environnement brutalement altéré, la sélection naturelle intervient et une accélération du processus évolutif tend à en résulter.
Les populations au sein de D. gallinae sont apparues en outre très différemment structurées entre faune sauvage et élevage. Alors que l’isolat sauvage de références IL offre des valeurs de polymorphisme équilibrées, des déséquilibres importants dans les élevages sont mises à jour dans les valeurs de polymorphisme (valeurs observées significativement inférieures aux valeurs attendues).
Plus précisément, des déséquilibres importants sont constatés chez tous les élevages sur la base de la Tropomyosine, mais seulement chez certains d’entre eux sur la base de la mt-Co1 (nombre d’haplotypes et diversité haplotypique très inférieurs aux valeurs estimées).
Toutes les populations d’élevages semblent avoir subi un événement fondateur ancien, dont la marque est importante sur les caractéristiques du polymorphisme des séquences de Tropomyosine, si l’on compare les isolats d’élevages et les isolats sauvages. Elles semblent en outre avoir une origine commune (poules pondeuses de France, du Danemark, de Belgique, de Pologne, poulets de Bresse). L’information associée aux déséquilibres du polymorphisme relevés dans l’analyse des séquences de mt-Co1 est un peu plus délicate à manipuler. En effet, l’image plus récente dessinée par cette séquence mitochondriale souligne des différences d’un autre ordre. L’isolat sauvage de référence offre une diversité mitochondriale très réduite (4 sites ségrégeants, 5 haplotypes), contrastant fort avec sa diversité nucléaire (37 sites ségrégeants, 19 haplotypes). Cela suggère que le génome mitochondrial de la population dont il est le représentant a eu le temps, depuis l’hybridation consécutive à la radiation évoquée ci-dessus, de dériver suffisamment pour retrouver un équilibre naturel en se développant à l’écart d’autres populations. Si l’on considère cet isolat comme le représentant de l’état « naturel » du génome de D. gallinae, les déséquilibres relevés ici dans les élevages, semblent représenter plutôt le signe de la mise en contact récente de populations longtemps isolées, plutôt que la marque d’une sélection brutale et passée d’un petit nombre d’individus. Or SK, représentant de la seule population dont nous connaissons l’histoire depuis 12 ans (élevage en laboratoire), manifeste un équilibre similaire à celui constaté chez IL sur la base de la mt-Co1, à la différence des autres isolats provenant directement d’élevages de pondeuse. Le confinement au laboratoire de cet isolat s’oppose à l’effet de l’introduction de populations différentes. Il est donc
129 apparu donc évident que les isolats testés prélevés dans des élevages de pondeuses français ont subi des mélanges récents de populations qui ont évolué assez longtemps séparément pour avoir développé des haplotypes de mt-Co1 très divergents. Paradoxalement, l’haplotype Co 1, extrêmement récurrent dans les isolats d’élevage de pondeuses français dont un petit nombre seulement d’individus a été séquencé, mais absent des élevages d’autre pays et des autres types d’élevage, vient corroborer cette interprétation : les moyens de transports motorisés utilisés au sein de la filière pondeuse, tout au moins en France, participent pour une grande part à la dissémination de l’acarien. Même si l’on ne sait pas exactement quel est (ou quels sont) les vecteur(s) de ces arthropodes pour transiter entre l’élevage et le véhicule motorisé (cartons, cages, charriots, poules, technicien, …), la forte décorrélation entre distances géographiques et caractérisation mitochondriale est un argument de poids en faveur du rôle primordial des mouvements commerciaux.
L’espèce D. gallinae se présente comme un complexe de lignées en ébullition, qui ont évolué séparément les unes des autres, puis se sont hybridées entre elles dans le passé. A l’heure actuelle, des mixages incessants sont en outre réalisés par l’action des transports commerciaux, rendant encore possible de nouvelles hybridations. D’une manière générale, une évolution accélérée et une flexibilité bien supérieure caractérisent l’espèce généraliste D. gallinae si l’on compare à ses sœurs spécialistes. Ces particularités intrinsèques associées à de nombreuses preuves d’une adaptabilité remarquable et couronnée de succès (spectre d’hôtes large, synanthropicité, expansion en cours constatée dans certaines zones du monde) clament le caractère fondamentalement apte à l’invasion de cette espèce. Le tableau réunit la plupart des caractéristiques recensées par Lee (2002), propres à ces rares espèces capables de passer toute les étapes de l’invasion décrites par Williamson (1996) : importation, introduction, établissement, développement nuisible. Et ce n’est pas un hasard si ce parasite généraliste se présente comme une entité dérivée dans un groupe de parasite dont l’état plésiomorphe de la spécificité d’hôte est l’état spécialiste. Une fois encore, le caractère spécialiste pour un parasite n’est pas un cul de sac évolutif (Desdevises et al. 2002) et l’augmentation de la spécificité d’hôte n’est pas nécessairement un progrès. Au sein de D. gallinae, la lignée L1 semble bien isolée des autres sur le plan reproducteur, quoique la divergence demeure encore très faible sur la base de la Tropomyosine. Elle pourrait représenter une espèce cryptique, mais il n’est pas certain que l’isolement ne soit pas réversible. D. longipes contient une espèce cryptique, déjà bien divergente sur la base de la Tropomyosine.
130 Publication V
Dermanyssus gallinae (Acari: Mesostigmata) possesses characteristics of an invasive species, compared to four other Dermanyssus species
Roy L.*, Lopes J.S.**, Dowling A.P.G.***, Chauve C.M.*, Buronfosse T.*
* Université de Lyon, Ecole Nationale Vétérinaire de Lyon, Laboratoire de Parasitologie, Marcy- L'Etoile, France ** School of Biological Sciences, University of Reading, Reading, United Kingdom *** Department of Entomolgy, University of Arkansas, Fayetteville, USA Corresponding author: Lise Roy, Ecole Nationale Veterinaire de Lyon, 1 avenue Bourgelat, 69280, Marcy l’Etoile, France. Tel: 334 78 87 25 25; Fax: 334 78 87 25 77; [email protected]
Running head: D. gallinae, invasive species
Abstract Some organisms may rapidly adapt to new conditions, whereas their close relatives may not, due to both various ecological parameters and intrinsic characteristics. Genetic traits allow some plant or animal species to become successfully invasive when an opportunity arises and host specificity within host-parasite systems is governed by similar rules. The Poultry Red Mite, Dermanyssus gallinae, is a common and harmful pest in layer farms. It is the only species of Dermanyssus present in farms and the least host specific species within this genus of bird ectoparasites. The aim of the study was to clarify the phylogenetic relationships of D. gallinae with four close relatives present in France and to test whether any conspicuous differences exist in the intraspecific molecular polymorphism between these species. A region of mitochondrial gene coding for cytochrome oxidase I and a region of the nuclear coding gene of Tropomyosin including an intronic part were tested using phylogenetic and population genetics tools. The phylogenetic relationships revealed a derived clade of D. gallinae+D. apodis preceded by D. carpathicus and the most basal species D. hirundinis and D. longipes Apparently, within D. gallinae, radiations have occurred a few times after the split between D. apodis and D. gallinae and subsequent interbreedings between nascent species interrupted most of these speciation events. A comparison of genetic data between wild isolates of D. gallinae, D. apodis and D. hirundinis shows a significant difference of haplotype number and diversity, the former having far more variable sequences than the two latters in both loci under test. Within D. gallinae, population structure clearly separates parasites from wild and domestic birds and there appears to be very little migration between the two host groups. Moreover, at least one ancient founder event is detectable in isolates from European layer farms. Recent migrations between French isolates show an important role of trade flow in the dissemination of D. gallinae between French layer farms. Present ecological and farm information along with obtained results led to several insights in favor of D. gallinae as a typical invasive species. Key words D. gallinae, phylogeny, population, invasive species, Dermanyssus
131 Publication V
Summary
Abstract ______131 Key words ______131 Summary______132 Introduction ______133 Material and methods______135 Biological material______135 Whole dataset______135 “Focused isolates” ______135 Farm 8006: exploration of mite movements within a farm building ______136 DNA extraction, amplification and sequencing ______136 Analyses ______136 Matrices ______136 Phylogenetics ______137 Haplotype networks ______137 Population genetics ______137 Results______141 Sequence data and haplotype characteristics ______141 COI______141 Tropomyosin ______141 Haplotype diversity and networks ______142 Consistency between mitochondrial and n analyses______143 Phylogenetic congruence at the interspecific level______144 ABC results ______145 Genetic relationships within and between populations ______146 Haplotype within focused isolates of D. gallinae and other species diversity and heterozygosity _____ 146 Interpopulation diversity and gene flow______147 Discussion______148 Phylogenetic considerations: a prolongation to Roy et al. (2009a) ______148 Relationships of D. apodis with other species of the genus Dermanyssus______148 Position of D. hirsutus ______149 Derivation of D. gallinae: radiation and subsequent hybridizations ______149 D. longipes ______151 Intraspecific levels ______152 Population structure ______153 Range of mite’s mobility ______156 Genetic variability according to the mite species______158 Conclusion ______160 Acknowledgments______161
132 Publication V
Introduction Along with various ecological parameters, some intrinsic characteristics underlie the potential for organisms to adapt to a new environment. Within a given zoological or botanical group of species, some entities may possess a greater ability to colonize a new habitat than others. This is the case for invasive species, whose close relative may not always be able to invade as diverse habitats. Lee (2002) established that invasive organisms are characterized by diverse characteristics, related to the genetic architecture of the species, which allows the species to rapidly adapt to a new environment when the opportunity presents itself. In a similar way, in host-parasite systems, the host range may be dictated partly by extrinsic ecological parameters and partly by intrinsic characteristics. Combes (2000) defined two filters genetically controlling parasite-host systems: (1) the genes or gene combinations involved in the encounter between the infective stages and potential hosts and (2) the genes or gene combinations implied in the post-encounter compatibility, i.e. the "durability" of the system. Thus, the host range may vary from one parasite species to another closely related one partly due to the physical possibilities of switching to various hosts (ecological parameters) and partly due to intrinsic characteristics allowing the parasite to rapidly adapt to a new host (ie. to a new environment). Generalist and specialist species in the Trematoda genus Lamellodiscus not only tend to respectively group together in phylogenetic topologies (Desdevises et al. 2002), but also generalist species have a much more important genetic (and morphological) diversity than specialists (Kaci-Chaouch et al. 2008). Authors of the latter article show that such a difference is rather an indicator of an intrinsically different genetic architecture allowing more variable species to adapt to a wider range of hosts/environments, than a consequence of the genetic isolation of the specialists. Finally, authors of the former paper also showed that specialists were not “dead-ends” as usually believed. Generalists in Lamellodiscus appear as derived (“more evolved”) species.
In contrast to typical parasites, micropredators provide more distant relationships with their individual hosts, as they sample food only as needed and typically do not continually stay on the host. Population structures in mosquitoes – whose adult females are micropredators - are mainly dependent on factors that are unrelated to their prey (macroenvironment), such as rainfall and opened water containers (Paupy et al. 2005), and reproductive isolation by geographical distance is very reduced (Gorrochotegui-Escalante et al. 2000). Following Kuris and Lafferty (2000), Dermanyssus species in the gallinae-group fall into the micropredator category, not in the typical parasite category since adult females feed successively on different host individuals like mosquitoes or bed bugs. Thus, within Dermanyssus, Moss (1978) highlights two different ways of life. Species in the hirsutus- group live almost permanently on a host, regularly taking small blood meals, which do not result in a conspicuously distended opisthosoma, and laying eggs among feathers as opposed to the nidicolous species of the gallinae-group. Many experiments performed on D. gallinae De Geer, 1778 (Wood 1917, Reynaud et al. 1997, Nordenfors et al. 1999) have shown that most of their time is spent within various crevices present around the bird and that only a short amount of time is needed to obtain a blood meal (1/2h-1h1/2 in D. gallinae). Life-cycle in the gallinae-group is as follows: egg is laid in the environment, larva moults into protonymph without feeding, protonymph and deutonymph each need one blood meal to moult into the subsequent stage (deutonymph and adult, respectively). In these species each blood meal is sizeable, resulting in engorgement and massive expansion of the opisthosoma (Radovsky 1994). During the adult stage, the male does not feed and the female needs only one blood meal per gonotrophic cycle (one egg laying session each and typically eight gonotrophic cycles per individual). Overall, feeding habits in adult females and several life history traits are strikingly comparable to micropredator bed bugs, much more than to
133 Publication V mosquitoes whose young stage ecology and winged condition represent important differences with Dermanyssus. Micropredator species within genus Dermanyssus were until recently considered to have a very broad host range. However, unexpectedly, some of the micropredator Dermanyssus species turned out to be more specific than previously believed (Roy et al. 2009b). As they are not winged, it is likely that these species are more dependent on the prey’s microenvironment, even if not necessarily on the host individual, and therefore constitute an intermediate case between winged micropredators and typical parasites for the purpose of studying characteristics associated with parasite dispersal. Dermanyssus gallinae is the only species of the genus currently infesting European poultry farms, mainly in layer farms, whereas four additional closely related species are commonly found in wild bird nests (Roy et al. 2009b). In these layer farms, D. gallinae causes problems such as downgraded eggs, stress, anaemia, and increased mortality, and is therefore a pest of great economic importance. This species has also been shown to inducing problems in wild avifauna (ex. pigeons, Clayton and Tompkins 1994, 1995). Other species encountered in wild avifauna in France are D. hirundinis Hermann, 1804; D. longipes Berlese and Trouessart, 1889; D. carpathicus Zeman, 1979 and D. apodis Roy, Dowling, Chauve and Buronfosse, 2009 (all in the Moss’ gallinae-group). Of these species, only the former’s impact on hosts has been studied and that was in the United States (Johnson and Albrecht, 1993; Pacejkaa et al. 1996, 1998). None of these studies resulted in an observed impact even though mite population growth was conspicuous. Within Moss’gallinae-group, Roy et al. (2009a,b), highlighted a strong opposition between two main clades, one of them consisting of D. gallinae haplotypes and potentially cryptic species (“synanthropic clade” or clade B in Roy et al. 2009), the second one grouping D. hirundinis, D. longipes, D. carpathicus and D. apodis (“tough clade” or clade A in Roy et al. 2009 due to their presence only in wild avifauna).. This opposition seemed to be correlated with a few ecological traits, and especially with the level of host specificity. The five species of Dermanyssus examined in Roy et al. (2009b) tend to group together, based on COI, according to their respective level of host specificity, the latter clade grouping specialist species, the former a diversity of generalist lineages. Such a distribution of host specificity levels had already been noted in the fish parasite genus Lamellodiscus (Trematoda: Monogenea), composed of typical parasites (Desdevises et al. 2002). And yet, these two parasitic taxa possess very divergent habits, the mites acting as micropredators and the trematodes as a more typical ectoparasite. As unwinged micropredators, Dermanyssus species are likely to be tightly bordered on the hosts’ microenvironment, and so may have an intermediate behaviour between typical ectoparasites and mobile micropredators. Although members of the gallinae-group are nidicolous, Roy et al. (2009b) established that the bird host was also a vector of the mites, but that prolonged contact with the new nest seemed necessary for invasion by the mite. This study has the following two objectives: (1) Assess the utility of a newly developed and original nuclear marker (Tropomyosin exon n, intron n and exon n+1) for the exploration of interrelationships between and within species belonging to Dermanyssus and comparing these results to previous studies (Roy et al. 2009a,b) (2) Address the following questions at an intraspecific level: a) Are the generalist Dermanyssus lineages effectively composed of cryptic species (potentially making them as specialized as the "tough species") b) Do generalists have more evolutionary flexibility than specialists? c) What are the respective roles of commercial movement and potential exchange between wild and domestic birds in the dissemination of D. gallinae populations?
134 Publication V
In order to sufficiently address both objectives, a combination of haplotype-based phylogenetic and genealogical approaches, tests based on coalescent theory, and statistical analyses of haplotype frequencies and diversity were conducted. Data from two independent loci, COI (mitochondrial) and Tropomyosin (nuclear) from isolates of various geographical, ecological and host origins were utilized in this study. The first objective will be dealt with by comparing obtained results from Tropomyosin with previous results utilizing COI, 16S, and ITS (Roy et al. 2009a, b). The second objective will be dealt with by the means of intraspecific considerations and of a recurrent linkage with previously known life history traits and newly observed ones. The inclusion of isolates of D. gallinae along with isolates of the four other French species will allow the comparison of obtained information on genetic diversity in the focused species and closely related species. Thus, these species are phylogenetically, but also ecologically close to D. gallinae, as all species tested are hematophagous ectoparasites and have nidicolous habits. Material and methods
Biological material The location, host species, mite species and other information linked with sampled mites are listed in Appendix 1. Mite isolates have been sampled from farms or from wild bird nests as described in Roy et al. (2009b). The distribution of samples is rather large and diverse within France, and includes a few samples from other European countries (ex. The Netherlands, Belgium, Poland, USA) (Appendix 1). Nests were analyzed using a method described by De Lillo (2001) except that no sodium hypochlorite was added to the water solution to wash the stack of sieves and that the sieves had a somewhat different mesh width (top to bottom: 2500 µm, 1400 µm, 180 µm, 100 µm).
Whole dataset One isolate corresponds to mites of a single Dermanyssus species, isolated from an individual nest or from a group of nests closely located to each other in a bird colony (wild avifauna) or from a single building (farms). From each population, 1-5 individuals have been separately sequenced.
“Focused isolates” From six of these populations (1 D. apodis, 4 D. gallinae s. str., 1 D. gallinae special lineage L1 ; see Roy et al. 2009a), 18-24 individuals have been separately sequenced, in order to get an overview of the intrapopulation variation: 4 populations in D. gallinae (SK, 8006B1, 8020, IL), 1 in D. apodis (GO). Moreover, 21 individuals belonging to D. hirundinis collected from barn swallows distributed around France were included in the analyses. Due to conspicuously weak variability in both genes tested, it is handled here as a pseudo isolate among focused isolates. These seven groups will be referred to as “focused isolates” and used specifically for population genetics analyses. The 4 isolates of D. gallinae s. str. were selected as follows: 3 farm populations and 1 wild population. Isolate SK was sampled in a Danish layer farm in 1997 and, since then, it has been cultured in lab by O. Kilpinen (Lyngby, Denmark), 8020 and 8006 directly come from two different layer farms (NW and Center of France, respectively) and IL was isolated in nests of a colony of European starlings nesting in a group of nest boxes installed and maintained by J. Komdeur near Groningen (The Netherlands). The L1 focused isolate was sampled in an amateur pigeon breeding facility in France (9001).
135 Publication V
Farm 8006: exploration of mite movements within a farm building Finally, 16 individuals randomly selected from a mixed aggregate of mites sampled from six different points in a single building (B5) in farm 8006 (see Appendix 1) were sequenced for COI. The obtained haplotype data were compared with COI obtained from a single point in a neighbouring building on same farm (focused isolate 8006B1).
DNA extraction, amplification and sequencing
DNA was extracted from individual adult females following a protocol that preserves an intact cuticle for voucher preparation and microscopic observation. Only females were sequenced in order to get maternal and paternal information from nuclear Tropomyosin as D. gallinae has been shown to be haplodiploid and reproduce through arrhenotokous parthenogenesis (Oliver 1966; Hutcheson and Oliver 1988). From most individuals a 700-800 bp amplicon of COI (total length dependent on primer pair used) and a 600-700 bp amplicon of Tropomyosin were separately isolated by PCR, and then sequenced. PCR was performed in either a Biometra TGradient or a MWG AG Biotech Primus 96plus thermal cycler in a solution containing 2 µl of template DNA, 2.5 units of Taq polymerase, 10 nmol of dNTPs, 20 pmol of each primer and 3.4 mM (COI) or 1.4 mM (Tropomyosin) MgCl2. After an initial denaturation step (95°C) for 10 minutes, 40 cycles of 20s at 95°C (denaturation), 30s at 52°C for COI or 56°C for Tropomyosin (hybridization), and 90s at 72°C (extension). A final extension step was carried out for 10 minutes at 72°C. Several primers designed for amplification of both DNA regions from various species are listed in Table 1 and were chosen to perform PCR under the same conditions for each of the two loci. Negative and positive controls were run with each round of amplification. PCR products were checked by electrophoresis in a 1% agarose gel. PCR products were sequenced by Genoscreen (France, Lille) using a 96-capillary sequencer ABI3730XL. For sequencing PCR primers were used and in some instances, in cases of Tropomyosin heterozygosity, internal primers were designed and used to effectively separate and sequence both alleles. Analyses Matrices DNA alignments were performed using MUSCLE 3.7. Without refinement, MUSCLE has been shown to achieve accuracy statistically indistinguishable from T-Coffee and MAFFT, but overall is the fastest of the tested methods for large numbers of sequences (Edgar 2004). Seaview 4.0 (Galtier et al. 1996) was used for DNA alignment handling and seven different matrices were generated of four different types: - Isolate DNA matrices are complete alignments of all obtained DNA sequences (one matrix per locus) (alignment ISOL_COI = the matrix containing all mt-Co1 haplotypes; ISOL_TRO1 = a first Tropomyosin alignment including alleles from 82 mite individuals which serves as the reference for molecular characterization (cf. §Sequence data and haplotype characteristics), ISOL_TRO2 = the definitive Tropomyosin alignment containing all obtained Tropomyosin alleles). - Haplotype matrices are alignments of haplotypes as individualized using DNAsp v5 (Rozas and Rozas 1995; Librado and Rozas 2009) (one matrix per locus) - A matrix of encoded In/Del has been elaborated by encoding as discrete characters the presence/ absence and, when present, polymorphism of inserts at points where gaps are noted in the alignment of the whole Tropomyosin sequences dataset. This
136 Publication V
insertions/deletions have been there encoded as if they were morphological or biochemical characters; see Appendix 2. - An Arlequin matrix of Tropomyosin genotypes and a matrix of COI haplotypes for each focused isolate and each species
Phylogenetics Each of the haplotype matrices and the matrix of encoded In/Del have been processed in phylogenetic analyses, including data on all isolates belonging to the five targeted Dermanyssus species. One individual of D. hirsutus and two outgroups were also included.
Analyses For analyses using the maximum parsimony (MP) criterion, indivividual molecular datasets were run using PAUP 4.0. For DNA alignments of Tropomyosin, gaps were successively treated as missing and as a fifth state.
For Bayesian analyses the Tropomyosin data set was run using MrBayes (Huelsenbeck and Ronquist 2001; Ronquist and Huelsenbeck 2003). A model of evolution was applied to the dataset as determined for each gene by MrModeltest (Nylander 2004) for Bayesian analyses. Analysis of Tropomyosin used the GTR+ ī+ i model, which was determined in MrModeltest using Akaike information criterion (Akaike 1974). Parameters within the model were not specified (or fixed) and MrBayes was left to estimate these independently from the data during analysis. Analyses in MrBayes included two independent runs, each consisting of four chains and 5,000,000 generations. Appropriate burnins were determined based on stationarity being reached through the use of Tracer v1.4 (Rambaut and Drummond 2007). Some coalescence ratios inferred from obtained topologies are presented in Table 2.
Haplotype networks In order to roughly represent interrelationships between and frequencies of haplotypes within D. gallinae and to take into account the potentially reticulate characteristics of these a priori intraspecific interrelationships, parsimony networks of haplotypes in this entity were produced using Network 4.510 (Bandelt et al. 1999) using the median-joining network algorithm and with a post-processioning MP analysis.
Population genetics The aim of using population genetics tools to analyse the data was to recover demographic parameters such as ancient and modern population sizes, values of gene flow and time of divergence between the populations. Both inter- and intraspecific analyses were performed. The latter were performed solely within the D. gallinae specie since this group was by far the best sampled one. These analysis were conduced on isolates from various habitats with two different group divisions: ecological categories (French layer farms, Non French layer farms, Wild birds, Non hen farms); by isolate (corresponding to geographical categories : Netherlands, Denmark, France SE, France NW, France Center). Finally a third study involving interspecific relations was performed. Although still not very widespread, the use of population genetics tools in phylogenetics studies has been steadily increasing (Miller et al. 2005; Hamilton et al. 2005). Population genetics analysis using Approximate Bayesian Computation (Beaumont et al. 2002) has been shown to be particularly useful in interspecies inferences, namely in colonization situations (Pascual et al. 2007), in speciation due to geographic expansion (Ferran et al. in submission), allopatric speciation (Hickerson et al. 2006) or in presence of several expansion events as in the humans out-of-Africa model (Fagundes et al. 2007). Advantages of using population genetics tools
137 Publication V come from the use of well sampled populations to estimate demographic parameters mostly disregarded by phylogenetic approaches. This last study involved mites belonging to the three best sampled populations each from a different species (D. gallinae, D. apodis, D. hirundinis). The test was performed on individuals from bird nests only. In these analyses, farm and other human managed habitat have been excluded as only D. gallinae is present in such conditions (Roy et al. 2009b). In all the three studies an Isolation with Migration (IM) model was presupposed. This population model assumes the existence of ancient populations with constant sizes that splits originating two other populations in a single instantaneous event. After this splitting there is the possibility of gene flow between the newly formed populations or between themselves and other existent populations (Wakeley, 1996). This model was chosen because it allows for the study of the branching history of the population tree while permitting to study patterns of gene flow between the assumed groups.
ABC methods Likelihoods for IM models can only be computed for relatively simple scenarios containing few parameters (Hey and Nielsen, 2007). In fact, likelihood functions can be practically impossible to solve analytically when dealing with complex demographic scenarios (Marjoram et al, 2009). Since ABC methods facilitate the comparison of alternative models marginal to the parameter values without the need for calculating likelihoods (Beaumont et al, 2002), their use to solve phylogeographic related problems has become of great interest (Hickerson et al., 2006; Fagundes et al, 2007; Legras et al. 2007). The standard ABC approach involves two steps (Beaumont et al, 2002): a rejection step and a regression adjustment and weighting step. The rejection step consists of accepting simulations whose summary statistics are close enough to the values of summary statistics computed from the observed dataset. To assess this closeness, a Euclidian distance is computed between the entire set of normalized summary statistics and the normalized summary statistics calculated from the data. A set of values of the parameters is accepted when its Euclidian distance is within a certain percentage of the closest points to the studied data as in the study by Beaumont (2008). The second step is a local linear regression adjustment that attempts to model the relationship between the parameter values and the summary statistics. We assume that in the vicinity of the target summary statistics the relation between parameters and summary statistics is close to linear. This is the reason why the linear regression is performed only for the accepted set of values of the parameters. The regression adjustment has been shown to allow for a better characterization of the space problem since more points can be accepted (Estoup et al, 2004). Also in this step, each accepted set of parameter values is given a weight between zero and one that declines quadratically until a defined distance from the studied data set as used by Hickerson and co-workers (2006).
To reduce heteroscedasticity in the regression, all demographic parameter values were transformed on a log scale. The transformed values of the parameters were adjusted one at a time using a general linear regression on the accepted points. Adjusted values were then back- transformed taking the exponential for all parameters, to express posterior densities on a normal scale (Beaumont et al, 2002; Estoup et al, 2004). The use of a transformation has also the advantage of minimizing the appearance of values outside the prior ranges after performing the linear-regression correction. Previous studies have indicated that under particular circumstances the logistic and related transformations can lead to biases in the posterior densities estimated in the vicinity of the prior boundaries (Ferran et al, in submission). To avoid this problem we choose a log transformation which still allows for points at the lower boundary to be retained within the support of the model. By using this
138 Publication V transformation some points are adjusted by the regression to be outside the upper boundary. These points were then discarded, a procedure that has been shown to give better estimation (Ferran et al, in submission).
A standard backward coalescent process was implemented (Hudson, 1990; Nordborg, 2003) to simulate genetic data. This data are obtained by adding mutations under an infinite sites model for sequence data (Kimura, 1969). Hamilton and co-workers (2005) suggest running several hundreds of thousands to millions of simulations, depending on the complexity of the underlying model. In our simulations 5,000,000 values of the summary statistics sets were generated and a tolerance į = 0.001 was used to give 5,000 points from which parameters were estimated. When performing model-choice between the suggested different scenarios either 5,000,000 or 10,000,000 points were simulated and a value for the tolerance į was used in order to obtain the same 5,000 closest points. We used the mode of the posterior distributions as a point estimate of the parameter. The credible intervals were calculated around the mode, following previous studies by Hamilton et al. (2005) and Beaumont (2008).
A program developed by Lopes and co-workers was used to simulate genetic data in an IM model for any number of modern populations (Lopes et al, in submission). The regression step was performed in the version 2.5.0 of the package R (Ihaka and Gentleman, 1996). A script developed by Beaumont was used to perform the step (makepd.r, www.rubic.rdg.ac.uk/~mab/). For all the posterior density estimation from the adjusted sample of parameter values we used the locfit function (Loader, 1996).
Comparison of scenarios using approximate Bayesian computation Two analysis regarding population structure of D. gallinae were performed using an ABC method. The first one concerned the ecological context (host types, case I) being the groups divided in: 1) French layer farms; 2) Non-French layer farms; 3) French non-hen layer farms; and 4) wild birds. The second study reflected geographical locations and was composed of 5 groups corresponding to true populations (case II): 1) 8006 – in the centre of France; 2) 8020 in the Northwest of France; 3) IL in Netherlands; 4) JBO in the Southeast of France; and 5) SK in Denmark. In both cases, in order to estimate the branching history of the groups, an IM model was assumed. The first study was composed by the listed 4 groups, which corresponds to 18 possible branching histories (Stone and Repka, 1998). However a first approach was performed to assess the presence or absence of migration in all the 18 possible topologies, this led then to a 36 models comparison. The second study involved 5 populations, which correspond to 180 possible topologies (Stone and Repka, 1998). Instead of analysing the 180 single categories the topologies were first grouped in 30 clusters according to the order of the populations branching scheme (see Appendix 3). After this first approach the model choice was performed on the branching histories from the most supported clusters. A third analyse concerning a inter-species situation was performed. Three populations belonging to different species were considered (case III): 1) IL – D. gallinae; 2) GO – D. apodis; and 3) D. hirundinis. This analysis aimed to disentangle the branching history of the 3 species. Absence of gene flow between the species was assumed. The model-selection steps were performed before estimating the final demographic historic parameters, which were done conditional to the most likely scenarios, i.e. the ones with a higher Bayesian probability. In all of the comparisons the prior probability of each scenario was set to be the same (i.e. uniform prior distribution). The posterior probability of each model was estimated by performing the rejection-step followed by a logistic regression (Beaumont, 2008).
139 Publication V
Beaumont (2008) indicated that it is possible to sample the model indicator (i.e. {1, 2,…, m}) for “m” models (M1, M2,…, Mm) from a prior and treat this as a categorical random variable, X, in the ABC simulations. We can then apply a categorical regression to estimate P(X=x1|S=s’), where x = 1, 2,…, m is the indicator for model Mx and s’ is the vector of the summary statistics that summarize our observed data. A scheme of weighting was also employed, with weights coming from the same Epanechnikov kernel, as in the standard regression procedure. The regression-step was performed using Beaumont’s R script calmod (http://www.rubic.rdg.ac.uk/~mab), which needs the VGAM package Yee and Wild, 1996). This procedure has been shown to substantially improve previous methods to select among different models using ABC (Fagundes, 2007; Beaumont, 2008).
Prior distributions of parameters The priors for the demographic parameters were chosen according to information available from the literature. In the absence of information broad priors were used so that all the realistic scenarios could be taken in account (Appendix 4). Mutation rates both for COI and Tropomyosin were treated as a nuisance parameter. Therefore, a broad prior was used for the loci mutation rates to account for the uncertainty on the estimates. When considering both loci at the same time the variation in mutation rate between them was accounted for by using large prior containing the ranges of values of the priors of the two loci taken separately. In a coalescent method the time is measured typically by generations. For this reasons the choice of generation times was important to translate divergence times in terms of years. The duration of the mites’ lifecycle is highly dependable on the presence or absence of hosts. This can vary from 9-15 generations per year in the wild to around 36-60 generations per year in farms, where hosts are fairly available. However, because we are working with values of time typically on the order of hundreds of thousands of years, whereas the domestication of hens as began at the most 8,000 years ago (Siegel et al., 1992; Yamashita et al., 1994), we assumed an average of 12 generations per year. Furthermore, the domestication of hens in Europe should have happen much later than in its place of origin (West and Zhou, 1988).
Choice of summary statistics The summary statistics were chosen according to their success in previous ABC studies (Beaumont, 2008). Three summary statistics were then calculated for each sampled deme: number of haplotypes, h; number of segregating sites, S; and the average number of pairwise differences, ʌ. These were computed for each of the populations taken individually and for each of the pairs of populations pooled together. Hence, the Euclidian distances were computed from a total of 30 normalized summary statistics in the first case with 4 groups and 45 normalized summary statistics in the second case with 5 populations considered.
Statistical analysis of haplotype frequencies and diversity Statistical analyses were performed between and within species and between and within the 6 focused isolates. They were performed using the isolate DNA matrices of COI haplotype sequences and individual Tropomyosin alleles (phased alleles in heterozygous individuals and duplicated homozygous sequences, in such a way that sequences represent the diploid state of chromosomes). Polymorphism in haplotype sequences (COI haplotypes and separated Tropomyosin alleles) within and among the seven “focused isolates” was examined (gaps excluded and as the fifth state in Tropomyosin) and migration rates were estimated using DnaSP v5 (Rozas and Rozas 1997). We estimated the number of segregating sites (S), the average number of nucleotide differences (k), the nucleotide diversity (S1), and the nucleotide diversity with Jukes and Cantor correction (S2). Pairwise genetic distances were computed using Fst (Hudson et al.
140 Publication V
1992) and statistical significance assessed after 1000 permutations in all cases using Arlequin 3.1 (Excoffier et al. 2005). Effective migration rates (Nm) were estimated from Fst and compared with results obtained via popABC. Additionally, in order to obtain an overview of observed polymorphism significance, simulations were performed using the coalescent process tool in DnaSP v5. Based on the coalescent process for a neutral infinite-sites model and assuming a large constant population size (Hudson 1990), allowed us to obtain empirical distributions for two of the above observed parameters with the confidence limit for a 95% interval from the number of segregating sites (S). This way, the expected number of haplotypes (h) and the expected haplotype diversity (Hd) were compared to corresponding observed data. Results
Sequence data and haplotype characteristics A 543 bp piece of COI from 211 individuals and a 615-694 bp piece of Tropomyosin from the same individuals was included into the analyses. The COI amplicon was also obtained in 41 additional individuals and the Tropomyosin amplicon in 16 other individuals.
COI The COI gene fragment corresponds to the positions 403-945 of the Varroa destructor COI gene and 135-315 of the V. destructor protein (Navajas et al. 2002) and is that used in Roy et al. (2009a and b). All sequences were free of stop codons and amino acids that have been noted highly conserved in insects (Lunt et al. 1996) were equal in all sequences. Based on Lunt et al. (1996), the analysed part of the protein comprised 181 amino acids with two complete and one partial external loops, two complete internal loops and four complete and one partial membrane spanning helices. The alignment was unambiguous and free of gaps.
Tropomyosin The nuclear Tropomyosin gene fragment involved in the whole analysis corresponds to 10 bp of exon n, a 585-664 bp intron n and 15 bp of exon n+1. Intron n is located between positions 551 and 552 of coding gene in Boophilus microplus, based on the complete CDS published in GenBank by C. Johnson (AF124514) and between positions 490 and 491 of the CDS sequence of D. gallinae published by Nisbet et al. 2006 (AM167555). In order to check the homology of aligned introns, larger Tropomyosin fragments from 1-2 individuals of four Dermanyssus species were first sequenced (individuals GO593, MAR1, 8004b, RQ18, JBO49DL2; see species and EMBL accession number in Appendix 1). This way, five sequences, including a 62-115 bp portion of exon n, the focused intron and a 53-80 bp portion of exon n+1 were aligned. This allowed confirmation of homology. Additionally, an alignment was performed with the above sequences after the intron was removed. The portions of coding region provided in the present study was exactly the same as the corresponding part in the D. gallinae CDS sequence in extended sequences of individuals of D. gallinae, D. carpathicus and D. longipes. In the sequences of the two individuals of D. apodis, a single nucleotide polymorphim in exon n and 1 in exon n+1 was noted (“C” instead of “T” at position 489 (exon n) and at position 498 (exon n+1) of Nisbet et al’s CDS). As for the translated amino acids sequences, they were free of stop codons and identical in all six Dermanyssus sequences and very close to B. microplus (differing by only 3 amino acids).
Within the introns, more than 50 sites involve insertions/deletions, but in many cases a fixed series of 3-5 bp (and even up to 15 bp) is inserted/deleted, resulting in inserted/deleted 35 bp- portions in the whole dataset of gallinae-group individuals (see Appendix 2). One region
141 Publication V involves some microsatellite motifs, whose number is strongly varying between species, between populations and within populations. Sites with insertions/deletions have been recorded based on alignment ISOL_TRO1 (available on request from LR) and their distribution all along the region under test is located only on intron n and is rather regular when the five focused species are included (Fig. 1). Anyway, the first hundred and the last hundred base pairs are free of insertions/deletions. When considering only populations of the species of economic interest, D. gallinae, 12 regions with insertions/deletions are observed (noted A, B1, B2, C1, C2, D, E, F, G, H, I, J). Region J is the site with varying number of microsatellite motif repetitions (mainly TGA). The inserted/deleted bp series in region G is polymorphic (CAGT /GAGC). The ten remaining inserted/deleted series have been noted as monomorphic in the present study. Note that inserted/deleted sites are found mainly in addition in D. gallinae and D. apodis (and in subtraction in D. longipes, D. hirundinis, D. carpathicus). As a result, sequences of D. apodis and D. gallinae populations are longer than in D. longipes, D. hirundinis, D. carpathicus (670-695 bp vs 615-652 bp).
Haplotype diversity and networks (Appendix 1 for all species and fig. 2 for D. gallinae) Fifty-five different COI haplotypes were isolated from the five species of Dermanyssus tested (three in D. longipes, three in D. hirundinis, six in D. carpathicus, four in D. apodis, 35 in D. gallinae). Sixty-one different Tropomyosin alleles were isolated from the five species (five in D. longipes, two in D. hirundinis, seven in D. carpathicus, two in D. apodis, 36 in D. gallinae. Pure synapomorphies, representing unique signs respectively common to all populations of a given entity (segregating sites, In/Del) were noted based on alignment ISOL_TRO1 and are shown in Fig. 3. Haplotype networks were calculated, using the Median-Joining algorithm, based on the whole data sets of COI and Tropomyosin sequences, including all sequenced individuals of the D. gallinae complex from every sampled isolate, and an outgroup. With H = 0, a treelike topology was obtained in both loci, without any loop between distant haplotypes, suggesting there was no important bias (such as recombination or recurrent mutation) (Bandelt et al 1999). The obtained haplotype networks are shown in fig. 2, with different colours/textures according to the six following categories: wild birds vs birds from farms (ex. French layers, other European layers, amateur layer, chickens, pet birds). The figure presents topologies obtained with epsilon = 10, in which numerous hypothetical haplotypes (corresponding to unsampled haplotypes) are present and form loops with close haplotypes in most cases. Only five Tropomyosin haplotypes have been sampled from different ecological categories (Tro_1, Tro_2, Tro_3, Tro_8 and Tro_11), other Tropomyosin and the whole COI haplotypes are restricted respectively to a single category. Isolates of L1 are grouped together in the networks for both loci, and their cluster is much more distant from the D. gallinae populations in COI-based than in Tropomyosin-based topology. This lineage emerges among other populations of D. gallinae. Isolates from wild avifauna provide a variety of haplotypes, as opposed to domestic fowl in Tropomyosin (25 different haplotypes in wild birds vs 14 in fowls). In contrast, COI revealed less haplotypes in wild avifauna than in domestic fowls (14 different haplotypes vs 22 in fowls). Based on the Tropomyosin network, wild haplotypes are dispersed all over the network. Additionally, four different points of initial contamination are revealed in hen farms. Of these five points, only four appear as derived (Tro_3+21, Tro_17+16) and the two others are basal to some haplotypes from wild avifauna (Tro_1, Tro_2).
142 Publication V
Consistency between mitochondrial and n analyses Based on COI, several strongly characterized lineages have already been revealed within D. gallinae (Roy et al. 2009a and b). Of them, two provided apparently strong divergence, suggestive of reproductive isolation and were named L1 and L2 in Roy et al. (2009a). Moreover, in the present study, one COI haplotype is revealed to be conspicuously recurrent in French layer farms Co_1. Indeed, 12 of the 13 French farms under test possess this haplotype (8002, 8003, 8006, 8009, 8010, 8011, 8018, 8019, 8020, F29, F38, F56) (Fig. 2 and Appendix 1). In the whole dataset, 28 of 29 individuals under test in both COI and Tropomyosin containing this haplotype also contain Tropomyosin allele Tro_1, Tro_2 or Tro_3. Furthermore, one particular lineage is revealed by Tropomyosin-based analyses: lineage L3 groups together populations from wild avifauna only (Tro 23, 25, 52, 53, 57, 58, 59, 60). Individuals containing these alleles possess COI haplotypes which group together in both BA and MP analyses, but also include four haplotypes found from a Danish layer farm (Co 11, 12, 13, 14). These populations are strongly characterised by insertion/deletion events (Fig. 4C, insertions/deletions only), but also by mutations (Fig. 4B, gaps as missing data). In any case, these correspondences within D. gallinae between mitochondrial and nuclear topologies, although dominant, were not recovered in all individuals, as expected at the intraspecific level (Fig. 4A).
Comparison mt-/nDNA analyses Because cytoplasmic DNA in most organisms is effectively haploid and maternally inherited, they have a genetically effective population size approximately four times smaller than that of nuclear loci (Birky et al. 1989). Therefore, evolution in mt-DNA is usually faster than in nuclear DNA. Due to differences in the effective population size between mitochondrial and nuclear DNA evolution, the monophyly of alleles is expected to appear more quickly in mt-DNA than in nuclear DNA (Palumbi et al. 2001). In the present case, the ratio is reduced to three times due to the haplodiploidy of D. gallinae. Thus, this species has been shown to be haplodiploid with diploid females evolving from fertilized eggs (Oliver et al. 1966; Hutcheson et al. 1988). As these authors also observed similar haplodiploidy in a closely related family (Macronyssidae), we assume here that other Dermanyssus species reproduce the same way. Although the three-times rule of Palumbi et al. (2001) is not applicable as such due to the haplodiploid condition, the simple comparison of the external branch length / internal branch length in mitochondrial monophyletic groups and of corresponding monophylies allows determining that Tropomyosin is already deeply structured (Table 2). In the COI gene tree, branch lengths between clades are much longer than intraspecific branch lengths in the “tough species”, but not in the “gallinae complex” (Roy et al. 2009b). All entities with a ration > 2 recovered monophyly in the nuclear topology. On the other hand, none of entities with a ratio < 1,5 recovered monophyly in the nuclear topology, except D. gallinae. Except for L1 (ration 7.3) and Lmt1 (ratio 90), only few nuclear loci were expected to be monophyletic among the lineages within D. gallinae (1.5% for Lmt2 and Lmt3). Topologies obtained from Tropomyosin confirmed that lineages of D. gallinae drawn from COI were only rarely recovered, except for L1, but with very low divergence (2 segregating sites). The three other major lineages, including Lmt1, were not recovered. The ABC analysis assumed a no migration model at the interspecific level (case III) and led to migration models intraspecifically (case I, II), which is consistent with phylogenetic topologies.
143 Publication V
Statistic population genetic tests revealed much more differentiated populations in COI than in Tropomyosin (Table 3), exactly as it might have been expected due to the reduced effective population size in mitochondrial DNA and as it was suggested by haplotype networks: Fst values based on Tropomyosin were sharply higher in interspecific pairs (0.74-0.98) than in intraspecific non L1 D. gallinae pairs (0.00-0.32). As for the special lineage L1 of D. gallinae, obtained Fst values in pairs involving this lineage along with a typical D. gallinae isolate were intermediate (0.44-0.65). In contrast, Fst values based on COI were roughly similar between each of the pairs under test (0.77-0.99), except for pairs of D. gallinae focused isolates involving the individuals of farm 8006B5 sampled in six different points of building B5 (see above ; Fst 0.40-0.58).
Phylogenetic congruence at the interspecific level First, in Tropomyosin, the numbers of pure synapomorphies (representing unique signs respectively common to all populations of a given entity, as opposed to the whole remaining dataset) are by far the most important in D. apodis and D. carpathicus (52, 49, fig. 3). Other specific entities group with no more than ten pure synapomorphies (Fig. 3). In present haplotype trees (Fig. 4), the biclade topoplogy described by Roy et al. (2009b) based on mitochondrial DNA, opposing the group of tough species (D. carpathicus+D. longipes+D. hirundinis+D. apodis+hirsutus-group) to the gallinae complex is not supported anymore by Tropomyosin-based topologies. The group of tough species (clade A in Roy et al. 2009b) is revealed to be paraphyletic. Thus, this results in a large basal polytomy (or in an unsupported basal grade) based on gap as missing data analyses (either MP or BA), with D. hirundinis and D. longipes ENVL as independent haplotypes, haplotypes for D. longipes PAS and D. carpathicus accordingly grouped and a three species clade D: (D. hirsutus (D. apodis + D. gallinae)). Here, D. apodis is a sister to the complex of D. gallinae lineages. When considering gap as the fifth state (only MP), the topology appears much more resolved, with strong support values, forming rather like a grade. Tough species of the gallinae group are successively distributed at the basis of clade D: (D. longipes EN (D. longipes PAS(D. hirundinis(D. carpathicus(clade D))))). A similar topology, although less resolved, is revealed by using a matrix of encoded insertions and deletions (Fig. 4C). As in Roy et al. (2009a), the relationships of D. apodis with other species are not the same in mitochondrial and nuclear analyses. Based on COI, it groups as a sister to (D. longipes+D. hirundinis+D. hirsutus+D. carpathicus). Based on the nuclear Tropomyosin, it appears grouped together with D. gallinae with strong support (100% bootstrap and 93-100%, rel. Bremer index in MP, 0.91 in BA Fig. 4D). Moreover, the number of pure synapomorphies in Tropomyosin opposing the entity (D. apodis+D. gallinae) to the other three species of the gallinae group is rather important (55, fig. 3). Finally, three ecological observations corroborate such a phylogenetic relation: (1) living mite individuals within D. apodis appear visually as active as D. gallinae at room temperature when stimulated by slightly breathing on them (in contrast to D. carpathicus individuals, which tend to stand motionless, even under breathing stimulations), (2) mite prevalence and density within nests of colony-living birds (swiflets for D. apodis, starlings for D. gallinae) was similar (up to 79% prevalence of D. apodis in GO population, with > 500 mites in many nests; up to 95% prevalence of D. gallinae in IL population with > 500 mites in many nests – other species around 30-40% prevalence in RSA), (2) both are proliferating in nests of birds exempt of hygiene against chick droppings (no removal of chick droppings as in tits, redstarts, etc.). As a result, of the five species under test, D. apodis revealed to be the closest species to D. gallinae. Therefore, "focused isolates" for statistical analyses include one D. apodis population, in order to compare population haplotype diversity between these
144 Publication V phylogenetically and ecologically close species. The pseudo focused isolate DhirF, on the other hand, represents a more distant species (D. hirundinis).
D. longipes seems to be composed of two different lineages. Isolates from Passer montanus (JBO180 and PAS, Appendix 1) provide a single haplotype in mitochondrial as well as in nuclear topologies, whereas populations ENVL083 and ENVL088 possess 2-4 haplotypes per locus. Phylogenetic analyses revealed that these haplotypes group respectively together and independently from each other in Tropomyosin analyses, whereas only COI topologies present these two lineages as monophyletic (Table 2 and fig. 4). As a result, lineages of D. hirundinis and D. longipes, which appear to be closely related to each other, form a polytomy in Tropomyosin analyses. Members in this complex, which is basal to the whole Dermanyssus species set under test , remain ungrouped in gap as missing data analyses (both MP and BA). Moreover, the observed number of haplotypes (h) and haplotype diversity (Hd) are much lower than expected according to the coalescence calculation based on the number of segregating sites S within D. longipes (Table 4).
Isolates of the special lineage L1 (Roy et al. 2009a) group together in both mitochondrial and nuclear analyses (both MP and BA). However, they do not group in an exactly congruent manner with both loci, as this lineage appears as a sister to the remaining D. gallinae populations in mt-DNA-based topologies, whereas it arises from within the D. gallinae complex in nuclear DNA-based topologies. They diverge by 10-13 % from other D. gallinae lineages in COI (1-2 % between each another within L1) and by only two mutational differences (no particular insertions/deletions) from some other D. gallinae isolates in Tropomyosin. They are more differentiated in Tropomyosin from D. gallinae focused isolates (9001 vs D. gallinae s. str. Fst 0.44-0.65; Table 3) than they are between each other (Fst 0.00- 0.32), but not as much as from other species (Fst 0.97 with D. apodis). As for the special lineage L2 (Roy et al. 2009a), individuals of isolate LB18 (the only isolate of L2 sequenced with both COI and Tropomyosin), that were characterised by strongly divergent mitochondrial haplotypes, did not reveal any particular Tropomyosin allele. They share an allele that is very frequent in hen farms (Tro_1; fig. 2B). Anyway, only two different individuals belonging to a single isolate have been sequenced in this lineage. Therefore, the monophyly of L2 using the nuclear gene has not been tested in present study. Note that the particular COI haplotypes found in L2 seem to be very scarce, as they were isolated in natura in three different nests only through the whole study published in Roy et al. (2009a and b) and the present one.
ABC results
The aim of the first analysis of the D. gallinae groups divided by host types is to assess presence or absence of migration. The results show a clear presence of migration irrespectible of the branching tree history between the groups (Fig. 5A, Table 5, Appendix 4A). The presence of migration in IM models can lead to poor inferences of divergence times between populations (Beaumont, 2008). This situation was observed in this analysis with the posterior distributions for times (Table 5) being very close to the chosen prior distributions (Appendix 4A). Nevertheless the analysis strongly supports one topology against all the other. The branching history (((French layer farms, Non-hen farms), Wild birds), Non-French layer farms) has a probability value of 90% (Appendix 4A). Although the posterior distributions for migration rates are quite wide, having large 95% credible intervals, there is a notourious difference between ancient migrations and migrations observed in the modern groups, in
145 Publication V particular the gene flow in ancient times seems to be substancially higher than the ones at the present. The second analysis referres to well defined D. gallinae populations. Each one of these populations is associated with a different geographical location: IL (Netherlands), JBO (France Sout-East), 8020 (France North-West), 8006 (France Center), SK (Denmark). In a first approach we try to identify the inherent branching tree history between the populations. Since there are 180 possible topologies in a tree with 5 populations, we grouped them in 30 clusters according to their branching scheme (Appendix 3). The analysis was then carried out firstly in these clusters. The results show that clusters D, Q and U have a cumulative probability of being the true underlined topology of nearly 95% (Appendix 4B). We then considered only the topologies represented in those clusters and run a model-choice analysis. The most supported topology by the data is ((8006, IL) ((8020, SK), JBO)) with a probability of 72% (Table 6, Fig. 5B). Despite the high probability value the referred topology is far from having a strong consensus. In fact, analysis carried out with COI and Tropomysin data taken seperatly fail to support the same topology (Results not shown). It is not a surprise then that the posterior distributions of divergence times are very wide and close to the assumed prior distributions. Still, a common result between the analysis carried out with the whole data set and with one locus at the time is to indicate the presence of migration in all the ancient populations. Most of the 95% credible intervals of the migration rates, though, include 0 (Table 6). The final ABC analyse concerns time divergences between the species D. gallinae, D. hirundis and D. apodis. This analysis supported strongly the division of D.apodis firstly at around 3.6 Mya followed by a more recent division between D.hirundis and D. gallinae. Once again there was no consensus between the trees obtained using the locus information separately, namely using the Tropomysin data, and the tree obtained using the whole data set. Such results are quite interesting, suggesting the existence of genomic related pressures conditioning differently both loci. The posterior distribution of the more recent divergence time is not very informative, having a posterior distribution close to the chosen prior. The posterior distributions for effective population sizes, however, are quite sharp. They point to values between 30,000 and 90,000 with the 95% credible interval not going much higher than 100,000. We should be aware nevertheless that these values refer to the effective population size and not the census size.
Genetic relationships within and between populations
Haplotype within focused isolates of D. gallinae and other species diversity and heterozygosity The average number of differences K in COI haplotypes for “focused isolates” (Table 4) is the most diverse in D. gallinae population from farm 8020 (K=4.3) and 8006B1 (K=3.6). In contrast, in D. gallinae isolates SK (from farm, but cultured in lab since >10 years) and IL and in D. apodis isolate GO, almost no within population diversity was noted (K=0.3, 1.1, 0.4 respectively). Dermanyssus hirundinis pseudo isolate DhirF shows a higher K (K=4.6), similar to the K of 8006B1. But it is important to keep in mind that this group of individuals was sampled from across France, as opposed to the five other focused isolates. Moreover, sample 8006B5 (six points in a single building in the same farm as 8006B1) provides the highest K (14.7). In contrast, the average number of differences K in Tropomyosin haplotypes for “focused isolates” is completely nul in D. hirundinis pseudo isolate DhirF, whereas K is important in farm D. gallinae isolates under test (13.9-17.3), and by far the most diverse in IL (32.5). Dermanyssus apodis isolate GO is the only population with relatively comparable diversity
146 Publication V between mitochondrial (0.4) and nuclear (4.6) genes. The intraspecific variation within D. apodis is not clearly estimable based on present dataset, as only one isolate was sequenced from a substantial number of individuals (South France). Anyway, the only other isolate involved here of D. apodis (MAR, Center France) provided the same haplotypes in both loci. Moreover, three individuals from Corsica provided a COI diverging by only 3-4 nucleotides from GO and MAR (ie 0.5-0.7 %; acc. no FN398146), not included in present analyses). These are sizeable insights of the low variability within D. apodis, which appears to be independent of geographical location. The observed number of haplotypes (h) in both loci (Table 4) is similar to the expected one calculated based on coalescent process from the number of segregating sites S in wild groups except D. longipes and D. hirundinis (D. gallinae focused isolate IL, D. apodis GO, D. carpathicus) with high P values. In contrast, farm isolates show expected h values much higher than the observed ones (D. gallinae focused isolate 8006B1, 8020, P values <0.002). Farm focused isolate SK is somewhat special, as it shows much higher expected Tropomyosin h than observed, but an almost equal h in COI. French isolates of D. hirundinis did not provide any segregating sites in Tropomyosin. The haplotype diversity Hd follows a similar scheme except that expected value in COI is a little bit higher than the observed one in D. apodis GO and that observed and expected values are similar in the farm isolate 8006B5 (whereas, as in other farm isolates, observed h value remains much lower than expected). Observed heterozygosity along with the number of Tropomyosin alleles per population are shown in Table 8. An increasing gradient in the number of alleles present in each of the “Focused isolates” is noticeable from SK to 8006B1, with 8020 as an intermediate. The wild isolate IL provides by far the most important number of alleles. The observed heterozygosity and allele combinations showed some more or less important departure from the Hardy- Weinberg equilibrium in focused isolates. Due to the known haplodiploidy by arrhenotokous parthenogenesis (Oliver 1966; Hutcheson and Oliver 1988) within at least D. gallinae, and likely within the whole gallinae group, the allelic distribution is very likely to be strongly unbalanced (male haploid, female diploid). All the more, the sex ratio in most samples appeared to be unbalanced as well (more females than male). Not to mention various other parameters susceptible to induce a deviation from the Hardy-Weinberg disequilibrium such as inbreeding, this knowledge led us to consider the departure to be expectable and, as a result, to not consider alleles as independent for testing the significance of population differentiation. Therefore, permutation tests of Fst estimates based on Tropomyosin used the genotype as the randomisation unit instead of the allele.
Interpopulation diversity and gene flow As already sensed by observing the haplotype networks, based on COI haplotypes (Table 3), Fst values indicate a strong differentiation between pairs of isolates within D. gallinae in all focused isolates pairs (0.70-0.92). An intermediate differentiation between pairs involving 8006B5 (0.40-0.58) and a stronger differentiation in pairs involving L1 (0.94-0.98) are to be noticed. In contrast, Fst values based on Tropomyosin alleles show very weak differentiation in pairs involving any layer farm focused isolates (0.00-0.11) and weak but somewhat higher in pairs involving the wild isolate IL (0.26-0.32). As for L1, pairs involving individuals in this lineage (9001) show a higher differentiation (0.44-0.65). Estimated migration rates Nm calculated from these Fst values are the highest between farm isolates independently of their geographical location (SK vs 8020: f, SK vs 8006B1: 4.19) and reduced between farm and wild isolates (based on Tropomyosin) and almost null in pairs involving L1 (both loci) (Tropomyosin: 0.26-0.64, COI: 0.009-0.031). Furthermore, the ABC analysis also suggest
147 Publication V important gene flow both by considering different groups according to the host types and by considering the different geographically located populations.
Discussion
Phylogenetic considerations: a prolongation to Roy et al. (2009a) Considering species boundaries within Dermanyssus (Roy et al. 2009), i.e. branching points, roughly, a congruence between species and an incongruence within species can be noticed. Moreover, coalescence ratios and subsequent predicted percent of monophyly in nuclear topologies inferred from the mitochondrial topology are roughly consistent with observed nuclear topologies. Finally, the absence of detectable gene flow between species is accompanied by visible gene flow within species, as most of lineages drawn from mitochondrial COI are not recovered in Tropomyosin topologies, as pairwise Fst values based on Tropomyosin are much higher between species than between populations and as popABC runs led to migration models when processed on intraspecific divisions. All that suggests that both independent loci under test are reliable for investigation of species in the gallinae group within Dermanyssus at the inter- and intraspecific levels.
Relationships of D. apodis with other species of the genus Dermanyssus The relationships between D. apodis and other species appear slightly different in mitochondrial and nuclear topologies. The species tree is often not identical to the gene tree (Nichols 2001), due to several potential causes. Reduced effective population size (Ne) in mitochondrial DNA compared to nuclear DNA often causes a higher mutation rate in mitochondrial DNA resulting in less resolved internal relationship in mitochondrial topologies than in nuclear topologies. As a result, in mitochondrial gene trees, the most recurrent bias is due to homoplasy, and inconsistencies in nuclear gene trees are due to the stochastic effects of lineage sorting (McCracken et al. 2005). Additionnally, interspecific hybridization in some cases may induce reticulation. In Roy et al. (2009a), a slight incongruence was already visible between mitochondrial and nuclear topologies, but the nuclear gene region used in this study was not variable enough to establish it firmly: in the present study, D. apodis appears definitely as a basal taxon to all others in mitochondrial analyses (as a sister to all other tough species or clade A) or as a sister to the distal gallinae complex in nuclear analyses. The choice of the more appropriate gene tree for relationships between species requires some attention. Some authors considered more appropriate mitochondrial markers for inferring phylogenies at the specific level (Moore 1995 and 1996; Michaux et al. 2002) but they were dealing with organisms with smaller number of generations per year (birds, rodents; D. gallinae in natura, around 15 gen/y, in farms, >200 gen/y), and thus with likely reduced mutation rates in both mitochondrial and nuclear genomes. McCracken et al. (2005) recommended a balanced and reasoning approach, taking into account both advantages and flaws due to different effective population size Ne and considering first whether independent gene trees are adequately resolved and then whether those trees are congruent with the species history. In present topologies, no contradiction with branching point nor in branching order may suggest any introgressive hybridization between the five specific entities under test. The only disagreement concerns outgroup rooting, which acts either among the whole species under test, leading to a biclade topology in mitochondrial trees, or at the base of the Dermanyssus clade, with a scale-shaped topology and members of the hirundinis group basally distributed to D. gallinae in nuclear topologies. Of course, some incomplete lineage sorting events might be responsible of such different gene trees. Anyway, almost no support is provided for basal
148 Publication V nodes in the mitochondrial biclade topology whereas supports for basal scale are superior in the nuclear toplology (MP, BA). And Tropomyosin seems to be almost exempt of homoplasies, according to consistency and retention indices for MP topologies ((CI=0.8238, RI=0.9320 with gaps as the 5th state; CI=0.8023, RI=0.9263 with gaps excluded). Not to mention the fact that the strict biclade topology is not recovered in mitochondrial MP consensus tree, in which members of the hirundinis group form a basal polytomy, suggestive of the scale-like nuclear tree. For these reasons and as resolution is much higher with present intronic nuclear gene region than with ITS, but not contradictory (Roy et al 2009a), the Tropomyosin topologies are considered the most accurate representative of the gallinae group species tree. Suggestion that D. apodis might be intermediate between the clade of tough and synanthropic mites was evoked in Roy et al. (2009a) based on some ecological and phylogenetic observations. This is confirmed here by Tropomyosin analyses. Dermanyssus apodis is parasitizing swifts, which are commonly living in towns, appears in Tropomyosin topologies as intermediate between species living in natura and more synanthropic lineages of the species of economic importance. Dermanyssus gallinae L1 is the only other city nesting taxon in present study, as it lives within pigeon nests, common nesting-sites concurrent of swifts (Nature Midi-Pyrénées 2001). From the ABC inter-specific analysis with D. apodis, D. gallinae and D. hirundis, D. apodis seems to take the roll as basal clade with poor support, though. The divergence between this and the other two species seems to be placed around 3.6 Mya. Again discordance has been observed between the analysis with mitochondrial DNA and the Tropomysin data.
Position of D. hirsutus The respective positions of D. hirsutus and D. quintus (Moss’hirsutus-group) remained unresolved in Roy et al. (2009a). Unfortunately, we were unable to obtain D. quintus Tropomyosin sequence. Anyway, the position of D. hirsutus in our Tropomyosin-based topologies is similar to the position obtained in ITS for D. quintus. And yet, this was very different from the position of D. hirsutus in the same study. In Roy et al. 2009a, the position of these two species of Moss’ hirsutus-group remained unclear precisely due to the surprising basal position of D. hirsutus in ITS only topologies. Effectively, this position was not supported by any node value: it appeared separated from D. quintus in BA, and closer to D. quintus in MP analysis, but with absolutely no support (basal nodes BPP 0.75, 0.85, 0.56 in BA, % bootstrap/Bremer relative index 34/0.29, 12/0.11 in MP), although with numerous and conspicuous morphological characters in common. Present Tropomyosin-based analyses seem to resolve these interrelationships. And so, the inconsistency of the opposition gallinae-group vs hirsutus-group evoked by Roy et al. (2009a) is confirmed here and Moss’ hirsutus-group is definitely arising from within the gallinae-group.
Derivation of D. gallinae: radiation and subsequent hybridizations The grade-like MP Tropomyosin topology suggests a more derived state in D. gallinae than in the other taxa. But the BA topology does not exhibit longer distances for this species. This analysis does take insertions/deletions into account, although the congruence of gap only topology with others (Fig. 4C, gap only, encoded following appendix 2) strongly suggests that these parts of Tropomyosin sequences contain important and consistent phylogenetic information. A similar topology was already suggested based on ITS sequences by Roy et al. (2009a), but the low amount of DNA divergence in this sequence did not provide strong support.
149 Publication V
This grade-like structure also indicates once more, as showed in Lamellodiscus by Desdevises et al. (2002), that the specialist condition does not appear as a “dead-end”. Dermanyssus gallinae, the generalist species, is, if not the more derived species, at least one of the more distal ones with D. apodis, whereas basal positions are occupied by specialist species. The monophyly of this entity is not doubtful here as it is recovered in all topologies, but this is not supported by many synapomorphies and branch lengths are amazingly short compared to other species such as D. apodis, D. carpathicus, D. hirsutus in both loci (Fig. 4B, F). This induced a low predicted percent of nuclear loci recovering this monophyly (15%, table 2), suggesting that the date of this coalescence occurred much later than coalescences for other species. But the distance of the coalescent node for D. gallinae to the common ancestor with the closest species (considering that the right arrangement is recovered by nuclear topologies) is also the shortest. Either the two loci under test in D. apodis, D. hirsutus and D. carpathicus have evolved faster than in D. gallinae, or the apparent low rate of evolution within the latter is rather an artifact due to multiple radiations and recurrent interbreeding. The latter alternative appears much more credible as: (1) The significantly highest number of nuclear segregating sites within D. gallinae isolates (12-37 with gaps as missing and 49-117 with gaps as the fifth state, see Table 4, Tropomyosin) as opposed to isolates or even pseudo isolates in others species (respectively D. apodis: 4 and 9, French D. hirundinis: 0 and 0, grouped French and American isolates of D. hirundinis: 8 and 10, D. carpathicus: 11 and 16) is suggestive of the assemblage of formerly highly divergent haplotypes, (2) The two lineages L1 and L3 are more or less completely isolated from other lineages within D. gallinae, and both are arising from within the species. (3) ) The presence of an important ancient gene flow is indicated by the ABC analysis (Table 5-6). Also, in these studies the topologies turned out to be hard to recover, which is usually caused by population structures more complex than the IM model assumed.
The weak coalescence ratios, as well as the very small number of "pure synapomorphies" in D. gallinae as compared to other species seem to result from one or several radiations close to the date of speciation between D. gallinae and D. apodis, in the former’s lineage (hence the very short branch length). The "focused isolate" of D. apodis is the only species providing any deviation between observed and expected number of haplotypes in Tropomyosin using the coalescent process in DnaSP (haplotype diversity seems to be well balanced), whereas it shows no deviation in COI. In the same time, the branch length from the closest dichotomy is the longest either in COI or in Tropomyosin Bayesian inferences. This suggests there might have been some hybridization between divergent lineages within D. apodis, which would have resulted in an accelerated evolution. A similar event might have occurred within it close relative D. gallinae. Seehausen et al (2004) showed that interspecific hybridizations might induce adaptive radiations. And yet, given obtained nuclear topologies, a radiation seems to have occurred soon after the speciation between this species and D. gallinae, but subsequent hybridizations between radiate lineages in the process of speciation appear to have stopped speciation in most cases as often reported (Seehausen 2007). This resulted in an apparently reduced rate of evolution (short branch lengths), which may be explained by the successive reintroduction of ancestral polymorphisms. Some remainders of these successive hybridization events and of the adaptive radiation may be found in L1 and L3. Lineage L3 is likely a decaying old lineage, which shows much gene flow and was sampled in a multi lineage area (starling nests…). On the other hand lineage L1 (Roy et al 2009a, b) seems to clearly represent an entity in the process of speciation. That might be due to interbreeding secondarily associated with a break- down of the linkage disequilibrium (Seehausen 2007), which might be a consequence of ecological modifications encountered in this lineage since there is evidence that hybridization facilitates major ecological transitions (Rogers and Bernatchez 2007). This hypothesis is
150 Publication V supported by the long branch length in COI, along with Tropomyosin branching. According to obtained phylogenetic topologies, the reproductive isolation of the lineage L1 appears to be confirmed by the similar grouping in both nuclear and mitochondrial analyses, but the position of their Tropomyosin haplotypes in phylogenetic topologies is not as a sister to the remaining D. gallinae lineages. Moreover, retained model in popABC (case I: ecological categories) also shows, based on COI, an insertion of L1 within D. gallinae (Fig. 5; Non hen farms). On the whole, clues in favour of a branching of L1 among typical D. gallinae lineages rather than basal to all them grouped together are numerous. This supports the radiation hypothesis. The number of “pure synapomorphies” is very low and pairwise Fst values comparing L1 to D. gallinae focused isolates are lower (0.44-0.65) than Fst values against the closest species D. apodis (0.97), but higher than Fst values between D. gallinae focused isolates (0.00-0.32) (table 3). This is an incipient speciation in process, confirming the postulate of a species complex evoked by Roy et al (2009b). The cryptic characteristic of this nascent species is consistent with this hypothesis. Usually, in cryptic species, speciation has occurred very early, with no time enough to get marked in morphological characters. Moreover, it has been mainly found in pigeons (appendix 1), which makes it a specialist as seems to be the ancestral condition within the gallinae group. Finally, the above described radiation and subsequent hybridization events could be correlated with the multiple origin of domesticated chickens evidenced by Liu et al (2006). Following these authors, not a single initial strain of the wild red jungle has been first domesticated and not in a single asian area. The nine strongly divergent clades they detected in hens suggest different origins from so distant regions as Yunnan, South and Southwest China … Different lineages might have mixed together long before chicken domestication for an undetermined reason, which would have resulted in an adaptive radiation. A subsequent hybridization might have followed…
In sum, the generalist and synanthropic D. gallinae seems to be a species composed of hybridized species. The lineage complex composing D. gallinae is very likely to be a case in point of such man-induced outcomes. Hybridization is more known to enhance adaptation to new environmental conditions in some plants (Schweitzer et al 2002; Rieseberg et al 2003) than in animals. However, several recent studies recently highlighted similar phenomena in some animals (Schwenk et al 2008). Arnold (2004) recorded a variety of animal cases of yield increase in farm animals and development and virulence increase crop pests and disease vectors due to natural hybridizations. This "anthropocentric" review clearly evidenced the recurrent phenomenon of increased adaptability subsequent to interspecific hybridization, in some human-shaped environments. Similar yield increase is visible within species, by hybridization between differentiated populations. For instance, Edmands et al (2005) conducted experiments in some arthropods, which resulted in similar conclusions. Finally, Seehausen et al (2007) explored the correlation between human-induced homogenization of environment and loss of biodiversity and highlighted the numerous cases of speciation reversals following hybridization between young species, between which prezygotic reproductive barriers are not yet fixed. Loss of environment heterogeneity often results from man activity, and in turn results in the contact of previously either geographically (species issued from allopatric speciation) or ecologically (species issued from sympatric speciation) separated.
D. longipes Splitting within D. hirundinis and D. longipes is re-estimated with rather three reproductively isolated lineages present rather than two, with our D. longipes to be divided into two divergent lineages: PAS and ENVL08 (sampled in two different host genera). The strong
151 Publication V difference between much lower observed h and Hd (table 4) values than expected in D. longipes (as opposed to other tough species) reflects important divergence within this group, which is unlikely due to a bottleneck, but likely to a natural separation by reproductive isolation. This was already suggested by a few mutations on the ITS sequence (acc. no FM179377 and AM903310; Roy et al. 2009a), but due to the overall very small number of differences within this sequence in Dermanyssus, the divergence was not conspicuous enough. Based on the newly developped nuclear gene marker Tropomyosin, the divergence is clearly confirmed. It is likely that lineages PAS and ENVL08 constitute two cryptic species. A problem is the fact that isolates of both lineages come respectively from the same locality, located ca 250 km apart from each other. Although no comparable level of divergence has been noted between geographically separated isolates within other Dermanyssus species under test, the potential effect of isolation by distance is questionable. Anyway one sequence in Brännström et al. (2008) provides us with a proof that the observed divergence between the two D. longipes lineages is not solely a geographical characteristic. Thus, these authors published two ITS sequences thought to be from D. gallinae, one from farms’populations, the other one from wild bird nests’ populations. The latter isolates revealed to belong to D. longipes (Roy et al. 2009a), and possess exactly the same ITS sequence as isolates ENVL083 and ENVL088, sampled from great and blue tit nests. The corresponding sequence in Brännström et al. (2008) was from mites isolated from nests of flycatchers and warblers sampled in Sweden. No sharp specific morphological differences have been evidenced based on our cuticle vouchers between both lineages PAS and ENVL08. These data are in favour of two cryptic species within current D. longipes. The first one has been sampled from the Eurasian Tree Sparrow Passer montanus near Avignon (France), whereas the second lineage was found in Tits (Parus spp) near Lyon (France), flycatchers (Ficedula spp) and warblers (Acrocephalus spp) in Sweden. The former has been sampled from the type host genus (Passer) and the type host region, leading to considerate it is the name bearing lineage. The latter will have to be described soon, once additional material has been obtained from both lineages, in order to have a clearer overview of their respective morphology.
Intraspecific levels The contrast between Tropomyosin which provides haplotypes sampled from various categories of habitat and COI with single category haplotypes was not unexpected in so differently transmitted loci (Fig. 2A). Thus, due to the biparental transmission of nuclear genes as opposed to the maternal origin of mitochondrial genes, it is likely that the Tropomyosin-based overview corresponds to a much more ancient state (three-fold effective population size if compared to COI) (cf. above). The difference in the evolution fastness between mtDNA and nDNA could have allowed populations to differenciate based on their mtNA, and not yet on nDNA in case of isolation subsequent to above described hybridisation events. As in many species and due to the above explained difference in nuclear and mitochondrial mutation rates, mitochondrial haplotypes appear monomorphic or with a few nucleotide differences in wild "focused isolates" (GO, IL) or other wild isolates (JBO, ROL). D. apodis and D. hirundinis may represent additional references of wild isolates. Of course, they provide much less variable haplotypes in both loci than D. gallinae. But the French population of D. hirundinis DhirF is not a “true” isolate (a single bird host species, but France-wide sampling), and yet, no diversity in Tropomyosin (h=1) vs moderate diversity in COI, higher than in true wild "focused isolates" D. gallinae IL (S=4, h=5) and D. apodis GO (S=3, h=4) from wild birds, was noted (S=15, h=4) (see Table 4). In contrast, the congruence between both loci in D. apodis true isolate GO (2 haplotypes in each locus) confirms the
152 Publication V effective faster evolution of COI in species under test, and so its enhanced power for characterizing populations from each another.
Population structure Within D. gallinae, populations seem differently structured depending on the ecological context, especially opposing wild and domestic isolates. The ABC analyses, assuming an IM model, for the case with different host types [farms vs wild birds – farms including amateur hen houses, diverse types of industrial farms (see Fig. 5A, Table 5 and Appendix 4A)] generates a strongly supported topology affirming the existence of fairly defined populations (Table 5), although their interrelationships are not clearly resolved. Polymorphism analyses (Fst) also converge to support a structure where wild and farm isolates are at least partly isolated from each another. Gene flow is visible among farm isolates at least based on Tropomyosin (Fst). The ABC analysis also supports the presence of gene flow (Appendix 4A), in particular ancient migrations (Table 5). Regarding the division in geographic locations (Fig.6B, Table 6, Appendix 4B), the structure is weakly recovered using the ABC method suggesting no real geographical structure. The lack of consensus between mtDNA and nDNA studies (Results not shown) might have led to different degrees of differentiation been reached according to the examined locus. A differentiation seems to be already sensible from Tropomyosin sequences at a large scale (involving different isolates of various geographical origins), in different environments. At a smaller geographical scale (at the isolate level, consisting of mites from nests of a single bird colony or a single farm building ; see §Whole dataset), no structure is revealed by Tropomyosin sequences, whereas COI seems to have already enough derived to give their own characters to mites living in the same small area, ie to reveal their monophyly as described by Palumbi et al (2001). In the same manner, the high number of COI haplotypes in the whole D. gallinae species (35) and their diversity (up to 9 % divergence excluding L1, up to 12 % including L1) contrasts with their homogenization within the wild isolate IL (S=3) and the lab isolate SK (S=4). And more interestingly, no geography-linked structure is evidenced,although isolates from very distant locations have been included.
History of exchanges between wild and domestic birds Initial(s) contamination(s) has (have) necessarily occurred from wild birds as poultry breeding is a man-made condition. According to the network of COI haplotypes (Fig. 2), in L1, the basal position of haplotype from wild pigeons to farm pigeons suggests that we sampled by chance both ancestral and derived haplotypes in this case. In hen farm haplotypes, no similar observation can be done. And yet, given the diversity in both loci across the species D. gallinae, a nd especially among and between wild isolates, it is likely that we were unable to enough largely sample to get a representation of ancestral populations. Several insights seem to confirm that present samples omitted populations close to the common ancestor in wild birds and that the closest one here is from farm. Based on Tropomyosin, it appears that some alleles that are omnipresent in farms (French as well as from other European origins) are also encountered in wild avifauna (Tro_1, Tro_2, Tro_3, Tro_16, Tr_17). This would be consistent with the necessarily wild initial origin of first contaminations in farms. The aim of the ABC analysis concerning host types of D. gallinae was to desintangle this underlined topology. However, given the wide 95% credible intervals for divergence times and migration rates the recover of the branching historie is far from consensual.
153 Publication V
Moreover, more than 14 other alleles that are predominant in wild avifauna (found in the Dutch focused isolate IL, and in additional wild isolates from France; appendix 1) have not been encountered in farms (ex. Tro_27, Tro_50, Tro_51; appendix 1). This also suggests that shared haplotypes between wild and farm isolates more likely result from more or less recent exchanges of mites between wild and domestic birds, and that exchanges are from fowl to wild birds. Additionally, the total absence of COI haplotypes shared between wild and domestic bird isolates reveals that no very recent exchange between wild and domestic birds has occurred (Fig. 2). The farm haplotype Co_1 is basal to a very close haplotype sampled (a single mutated site), from a wild bird (Co_23), suggesting there might have been some relatively recent transfer from farms to wild avifauna, but if so, very uncommon since only a single individual case is present in the dataset. This uncommon transfer should have occurred at least after 6000BC (domestication of hen in China according to West and Zhou 1988), but more likely after 3000BC (first dispersion events of domesticated chickens in Europe, ibid.), which is not contradictory with the single mutated site. No other closely related haplotypes revealed any other exchanges between wild and domestic bird in non L1 D. gallinae populations. Finally, only one isolate from wild avifauna provided exactly the haplotype Co_1 (isolate Percno, in Roy et al. 2009b). This case should not suggest a possibility of fowl contamination by wild bird, but may simply result from an accidental transfer since concerned bird was Neophron percnopterus, a protected bird of prey that was cared for by some ornithologist societies (LPO) and precisely allowed to feed on dead hens directly collected from farms in the sampling area (France, Ardèche).
Founder event in farms and intra fowl industry mite dissemination
A very different population structure is revealed within D. gallinae isolates sampled in wild avifauna and isolates sampled in farms. The isolate IL taken as a feral reference provides balanced polymorphism values according to coalescent simulations using DnaSP, whereas important deviations between observed and expected values are revealed in farm isolates (observed values significantly lower to expected values). More in details, important deviations are visible based on Tropomyosin sequences in all layer farm focused isolates, whereas deviations characterize all layer farm focused isolates except SK based on COI sequences. On the whole, the possible interpretation for such departures leads to an alternative: the strongly reduced number of haplotypes and values of haplotype diversity may result from a bottlenecked population followed by an expansion (founder event) or simply give evidence of a recent admixing of long separated populations. Due to the fundamental difference in the mutation rate between both loci, it is assumed that Tropomyosin witnesses much older events than do COI. And yet, all farm populations seem to have undergone a common ancient event. In contrast, only French farm isolates appear to have been subjected to some more recent event, since the polymorphism of COI sequences within the Danish farm isolate SK looks like well balanced. The fact that all farm isolates show a similar deviation, with a strongly reduced number of haplotypes compared to their respective expected values and to the feral reference IL based on the number of segregating sites is strongly suggestive of a founder event consecutive to farming practices (use of pesticides…).
And it is likely that similar founder events are to be found in chicken French farms, as well as in non French layer farms, but not enough individuals/isolates have been sequenced to establish it with any statistical significance. Both isolates from chicken farms (8012 and BER)
154 Publication V group together based on COI haplotypes (Co_7 and C_15, relatively close to each other; fig. 2). Additionally, they share same Tropomyosin haplotypes as samples from French layer farms and from other European layer farms, suggesting that the former founder event has occurred before the separation of this population from French layer farms’ population. Moreover, demographic information obtained using case I and both loci could help leading to this interpretation. Indeed, the Ne of wild category (453000) along with internal nodes ancestral Ne (217000 and 264000) appear to be much larger than in layer farms especially in French farms (Non French farms: 24000; French farms: 2000), although the 95% credible interval expressed by quantile values are quite wide (Table 7).
The more recent picture which is drawn from COI sequences seems to be correlated to a very different event. Thus, the feral reference isolate IL shows a mitochondrial diversity very reduced (4 segregating sites, 5 haplotypes) if compared to its own nuclear diversity (37 segregating sites, 19 haploptypes). This suggests that the mitochondrial genome of the population it represents had time enough, since above described hybridization events, to derive and differenciate from each others and so recover a natural polymorphism balance. When considering this isolate as the representative of the “natural” state of D. gallinae mitochondrial and nuclear gene diversity, deviations noted here in farm isolates give evidence much more likely of very recent, if not contemporaneous, intermingling of separated populations than of bottlenecked populations. Indeed, the isolate SK is the only population of which pesticide exposition and isolation history is known since 12 years (cultured in laboratory, without any pesticide treatment and strictly confined). And yet, it shows a balanced polymorphism comparable to the isolate IL’s based on mitochondrial sequences, in contrast to field sampled farm isolates. The laboratory confinement of isolate SK keeps it from getting mixed with other layer farm populations. As SK was sampled from not a organically controlled layer farm, a potential pesticide-induced bottleneck explaining the deviation within directly field sampled isolates would have occurred less than 12 year ago or SK would have been able to enough derivate to get homogenized COI sequenced. A duration of 12 years spent within farms or laboratory for populations of D. gallinae might have allowed ca 400 generations to get completed, as mites in this species have shown being able to perform one individual complete development (from egg laid to egg laying) within 6-11 days in farm conditions (Tucci et al 2008) and as fowls are almost always available in layer farms (flock duration = 12 months, empty period between two flocks = no more than 2 months). Clearly, up to 5-6% divergence is not possible to get reached within 400 generations. Therefore, it seems impossible to correlate the COI deviations with a farming practice founder effect. Finally, mitochondrial information here gives strong evidence of recent or contemporaneous population exchanges and intermingling within French layer industry. Trade flows apparently play an important role in the dissemination of populations of D. gallinae at least within the French industry network. And the clearly uncorrelated geographical distances and mitochondrial characterization within French layer farm isolates, although wild isolates highlighted the natural isolation mark of COI sequences, are an additional clue in favour of this interpretation. For instance, the conspisuously recurrent haplotype Co 1 (along with close derived Co_5 and Co_10) in layer farms was encountered France-wide and in various types of industrial layer farms (see Appendix 1). As a result, motorized vehicles are obviously responsible of mite exchanges between farms, althrough we do not exactly know in details which precise vector(s) is (are) used by mites for transferring from farm to vehicle (boxes, cages, hens, men…).
Farm 8006 confirms the potential role of pullets as vectors for poultry red mites between farms: the multi-origins of pullets is to be correlated to the much lower haplotype diversity
155 Publication V
Hd as expected whereas the average number of differences K is the highest (Table 4) in sample 8006B5. With the apparent contradiction of a low Hd vs the highest K (14.7), this isolate is a case in hand of the very recent admixing of separated lineages within French layer industry. The layer farmer of 8006 is provided with young hens for layer flocks by more than 10 different hen breeders (one per flock, 10 different buildings with an overlapping turnover of flocks) distributed France wide. Sample 8006B5 is representative of the whole mite population of a farm building that is successively introduced to newly entering pullets from a diversity of hen breeding facilities.
Range of mite’s mobility Among the many possible ways of dissemination, human activity has been evoked above on a large geographical scale (France, Europe) or, on a smaller scale, between nests separated by several hundred meters (isolate IL). But at still smaller scales, within a farm building or at the nest level, within a precise site and excluding transport by another organism, almost no knowledge is currently available. Concerning the self movement of mites on a smaller defined volume, Roy et al. (2009b) reported an amazing case. In a barn housing chicken cages together with barn swallow nests, lots of D. gallinae individuals were found within and around chicken cages, some D. hirundinis individuals in swallow nests, but no cases of cross contamination. Especially, swallow nests were absolutely exempt of any D. gallinae individual, although located only two meters above chicken cages. The chicken farming schedule along with swallow chicks examination allowed deducing that D. gallinae mites were present at the beginning of swallow nesting in two nests and in the absence of chickens. The complete absence of even dead D. gallinae mites in hirundinid nests was surprising, as mites in this species are (1) quick and nimble runners (Clayton and Thompkins 1994, LR pers. observation), (2) able to switch instantaneously from galliform to passeriform birds and easily develop (Roy et al. 2009b), (3) have been recorded – at least LB18 (appendix 1) – in swallow nests. Moreover, since then, authors noted a second similar case (isolate TB08, appendix 1)). This might be due, more than to motion ability, to the ability of detecting host presence. Kilpinen (2001) has shown that temperature variations are one of the main stimuli allowing D. gallinae mites to find its host. This author evidenced that a gradient of temperature as subtle as 0.005°C/s is effective in activating D. gallinae mites. Other stimuli detected by this species are CO2 and vibrations (Kilpinen 2005). Additionally, some bird skin compounds may act as kairomones stimulating feeding behaviour (Zeman 1988). But experiments in these studies were performed in areas much more restricted than any barn. Thus, mites belonging to D. gallinae are known to be highly sensitive to some physico- chemical stimuli, but we do not know exactly within which range of distance, and as a result, cannot assess whether, in the present case in hand, mites in the vicinity of empty chicken cages did or did not detect swallow individuals only two meters away. Additionally, the sampling strategy in farm 8006 is bringing more insight into the reduced movement of D. gallinae. Indeed, in this free-range farm, COI sequence copies isolated from mites randomly selected from 6 different points within a single building (8006B5, points located 6-10 meters apart from each another) revealed significantly more diversity than in the corresponding single point samples 8006B1 (radius ca 30 cm). This strongly suggests that mites do not move a lot within layer buildings, subpopulations keeping roughly unmixed within a building. Although D. gallinae mites are physically able to run actively and go from one neighbor nest box to another as noted by Clayton and Tompkins (1994), they do not do so necessarily very often. The omnipresence of hens in a farm building during flock likely justifies the fact that mites do not need to move, although hens are moving themselves. Farm 8006 is not a cage farm, but a free range farm and sampling points were located at the connexion between roosts. As micropredators, mite individuals do not need to feed
156 Publication V successively on a single host individual. The roost connexions were, as usually, full of mite aggregates, which did not need to move due to the recurrent availability of host in these areas. Mullens et al. (2001) evidenced in the distantly related species O. sylviarum a gradual contamination of hen cages from one neighbor to another. Infestations increased first on the nearest hens and were detected on more distant hens only after mites had increased to moderate-to-high populations on closer hens. Of course, O. sylviarum possesses differences habits (staying and laying eggs on host) and life history (deutonymph does not feed), but its hematophagous habits are comparable to D. gallinae’s. And yet, a similar scheme is suggested by Clayton and Tompkins (1994) in D. gallinae following an experiment on pigeons involving a batch of nest boxes. Apparently, D. gallinae mites are unlikely to run from one nest to another above several meters distance (a range to be defined by some field/lab experiments). Above this undefined distance, it seems that a carrier is needed for mite dispersal (bird, man, trolley, etc.).
D. hirundinis is likely to disperse the same way. In Roy et al. (2009b), this species appeared to be specific to hirundinids in France (ten isolates), whereas the three isolates coming from the same site in the USA were isolated from nests of three different families of Passeriformes : tits (Paridae), wrens (Certhiidae) and swallows (Hirundinidae). Over the marked genetic divergence between American and French isolates, the different host specificity was explained by fundamental ecological differences between American hirundinids (Tree Swallow, Tachycineta bicolor) and French hirundinids (Barn Swallow, Hirundo rustica, and Common House Martin, Delichon urbica). The former is a cavity nesting bird, whereas the latter are birds that build their own nests. Therefore, the American Tree Swallow is commonly sharing nestboxes with House Wrens or with Tits, in contrast to French mason swallows, which never nest in wren or tit nest boxes. This explains the absence of D. hirundinis from the numerous tit nests analysed in previous and present study (around 200). Anyway, the House Sparrow (Passer domesticus) as well as the Wren (Troglodytes troglodytes) are known to reuse old nests of Barn Swallows. And yet, since then, three dead individuals have been isolated from wren nest (isolate LC083; appendix 1), belonging to the French lineages (haplotypes Co_ ?? and Tro_44). The question here is, are they individuals from the previous season, which would have developed on swallows and would not have parasitized wrens? Another clue in favor of a host spectrum directly correlated with nest- sharing and carriage by the bird host is the individual MG1 (appendix 1): an adult male isolated from litter of a young House Sparrow which had been taken in after falling from the nest. So far, only 16S and COI amplicons have been obtained (respectively 94-96% and 96- 99% identity with other French individuals). Both these elements suggest that D. hirundinis may switch from one host to another in case of nest sharing (from one year to another), as does D. gallinae. This confirms what Roy et al. (2009b) partly evidenced by performing some lab bioassays (some D. hirundinis individuals sampled from a swallow nest fed on canaries and normally developed). As a result, the observed reduced host spectrum in French D. hirundinis does not seems to be due to intrinsic factors, but much more likely to ecological factors, linked to bird host habits, as do American lineages in the same species. The filter one in Combes (2000) may be similar in D. gallinae and D. hirundinis.
Nevertheless, at least the French lineage possesses adaptability to a new habitat (Combes’ filter 2) different than in D. gallinae. Thus, although found on a House Sparrow, it has never been found in tit nests (a significant number of tit nests analysed), nor in ay other bird under test. And yet, the House Sparrow is not only known to reuse swallow nests (Weisheit and Creighton 1989, but also to use nestboxes made for cavity-nesting birds). If D. hirundinis is
157 Publication V able to feed on another passeriform species than the initial host and subsequently develop in case of bird nesting within the initial host nest, it does not seem to be able to develop in a – even slighlty - different habitat. Is the difference in nest arthropodofauna the cause? Are the components of the nest a cause (earth vs herbs)? It is not unlikely that once transferred to a Sparrow nest by the bird, the mite does not find any appropriate conditions for its development (no species of Dermanyssus detected in 12 nests of House Sparrow sampled in France analyzed and three nests containing D. longipes lineage PAS out of 17 nests of Eurasian Tree Sparrow analyzed).
And interestingly, Clayton and Tompkins (1994) deduced from their experiments involving D. gallinae along with some Mallophaga on pigeons and comparing their differential virulence that the virulence of ectoparasites is proportional to the amount of horizontal transmission. Ectoparasites which are able to disperse independently of hosts are extremely virulent, in contrast to those which are dependent on the direct contact with the host, which are not virulent. And yet, the impact of D. gallinae and D. prognephilus in wild avifauna has been shown with rather strong statistical significance (Clayton and Tompkins 1994; Moss and Camin 1970), whereas none of three studies on D. hirundinis (USA) has been able to evidence any impact on wrens (Johnson and Albrecht, 1993; Pacejkaa et al. 1996, 1998). Is this one additionnal clue in favor of the absolute need of nest sharing for transfer in D. hirundinis ? Even if D. gallinae seems to disperse the same way, its apparently increased flexibility (generalist condition) makes its transfer less insecure than D. hirundinis’, making it less dependent on its host. Note that phylogenetically close D. gallinae and D. apodis share a nimble mobility as well as high levels of prevalence within some nests in colonies, in contrast to D. carpathicus (pers. obs.). And yet, only these two species of the five under test were found directly on the host in previous studies (Roy et al. 2009b), and always as adult females, which might represent the stage/sex appropriate to dissemination (Roy et al 2009b). Maybe these species are simply more able to cling to their host by being more nimble than the others?
Genetic variability according to the mite species Present study provides several clues for assessment of a potential invasive character in D. gallinae. Lee (2002) established that successful invasive species in many cases: - possess either increased additive genetic variance (AGV) or increased epistatic genetic variance if compared to other species under test. - are favored by hybridization, along with multiple introduction events. Several invasive plants got established through multiple introductions, followed by hybridizations, which allowed the organisms to benefit from new AGV or new epistasis (joint effect of different loci). - are accompanied by rapid speciation. Indeed, invasions constituting frequently rapid evolutionary events, rapid speciation following invasions could occur through neutral and selective processes (Lee, 2002). Several clues in favor of the presence of a particular genetic architecture in D. gallinae are available in present data.
Different genetic diversities D. gallinae provides a much higher genetic diversity than the four other species under test (S, h, Hd, K, table 4). The higher genetic diversity of D. gallinae suggests this species may possess either increased additive genetic variance (AGV) or increased epistatic genetic variance if compared to other species under test. This is one of the characteristics in the genetic architecture of many invasive species (Lee 2002). Of course, the genetic architecture
158 Publication V was not strictly observed in present study, as our simple haplotype-based approach does not allow distinguishing between genetic drift and natural selection. Anyway, as suggested by Lee (2002), genetic diversity may be an indicator of genetic variance.
Differences are intrinsic Apparently, an important intermingling involving a large number of a diversity of haplotypes in D. gallinae is noticeable within a colony of starlings (population IL, 19 Tropomyosin haplotypes). But this does not seem to be solely correlated to the bird’s ecology, as we noted the exact contrary in D. apodis from a colony of swiftlets (two Tropomyosin haplotypes) and in D. hirundinis from 7 separated French colonies of barn swallows (one Tropomyosin haplotype). Thus starlings, swiftlets and swallows are used to reusing nests of other pairs in the same colony from one year to another (O. Caparros, CRBPO, MNHN, pers. comm.). This could suggest that D. gallinae is able to move from one nest to another by the mean of the bird host (see above). But the ability of D. apodis to get transferred by the mean of birds has also been shown by several adult females directly sampled on (flying) hosts (Roy et al. 2009a, b). The conspicuous stability of both genes under test in D. apodis and D. hirundinis strongly contrasts with their variability in D. gallinae and suggests that these species are intrinsically very different. The variability within D. carpathicus and D. longipes is difficult to be estimated due to the too weak number of sequenced individuals/isolates.
Incipient speciations and hybridization within D. gallinae Radiate nascent species apparently have inter-hybridized, which stopped the speciation process, as shown above. The subsequent diversity of alleles found in wild avifauna, as opposed to the very low number of alleles found in close species could have alleviated loss of AGV during founder events detected in the present study, and have generated novel genotypes, more adapted to their farm environment. Invasions constituting frequently rapid evolutionary events and rapid speciation following invasions could occur through neutral and selective processes (Lee 2002). And yet, several incipient speciations seem to have occurred within D. gallinae, one maybe being definitive (L1), some others having been aborted (hybridization; e.g. lineage L3). D. gallinae revealed here to be a species complex in evolution: the high variability of mitochondrial and nuclear haplotypes is correlated in one case with reproductive isolation according to phylogenetic analyses (L1) and in other cases, strongly divergent Tropomyosin alleles are mixed in single isolates with homogenous COI sequences, suggesting a relatively ancient intermingling between long separated nuclear alleles. Some entities have likely been in the process of speciation (ex/ L3), but most of these incipient speciations were not definitive. The L1 lineage is an example of a youngest species which is maybe in the process of becoming definitely isolated, but maybe it is still able to reproduce with other lineages of D. gallinae in case of opportunity. Some more recent events, linked with the modern commercial exchanges and other man- made environmental defragmentations currently allows some more recently isolated lineages (possibly nascent species) to remix with others and once more enrich the genetic diversity in this species. The very recent admixing of populations in farm isolates revealed by polymorphism analyses in COI at least predicts potential new hybridizations.
Increase of genetic diversity in D. gallinae has already proved being successful in increasing adaptability to a new environment Several insights into invasive characteristics in D. gallinae allowing this species to be a generalist have been revealed here. But also several data confirm that D. gallinae has already proved to be successful in adapting to a wide range of new environments.
159 Publication V
Desdevises et al (2002a) have shown that generalist species tend to group together in the phylogeny of genus Lamellodiscus (Trematoda: Monogenea), suggesting that the level of host specificity is related to some intrinsic characteristics. In the same point of view, Kaci- Chaouch et al (2008) has shown that genetic and morphometric variability within Lamellodiscus was correlated to the level of host specificity. These authors, after having demonstrated that increased variability in generalists was not a consequence of the large host range (natural selection in various habitats), conclude that, in Lamellodiscus, successful host switching is more likely to occur in parasite species exhibiting more intraspecific variability. As a result, the genetic architecture is different between generalist and specialist species within a given parasite taxon. And yet, within Dermanyssus less genetically variable (h, Hd, K) species are effectively more specific: D. apodis is only known from genus Apus and D. hirundinis is encountered in France almost exclusively on hirundinids and seems not to be able to quickly adapt to different bird nests (see above), in contrast to D. gallinae which has been noted in birds belonging to nine different orders and having very diverse habits (Roy et al 2009b). D. longipes PAS seems to be restricted to the genus Passer (Berlese and Trouessart 1889, previous studies of authors). Looking like intermediate species, D. longipes ENVL08 and D. carpathicus are known from two different bird families (Paridae and Muscicapidae), both in Passeriformes, but datasets in the present study are not sufficient to establish any correlation between the level of host specificity and the degree of genetic variability. The mono-haplotype characteristic of populations PAS and JBO108 in both loci under test contrasts with the two and four haplotypes found respectively in COI and Tropomyosin in populations ENVL083 and ENVL088.
Of the five species of the Dermanyssus under test here, only one is a true generalist, which does not allow consistent tests that phylogenetic groupings are correlated with the level of specificity as found in other groups by Desdevises et al (2002). Anyway, the strongly divergent variability noted in both loci within D. gallinae contrasting with related species strongly suggests that a different genetic structure has allowed this species to successfully colonize the recorded diversity of hosts, possessing a diversity of habits (nest hygiene…), and in a variety of habitats (in natura, cities, farms…).
Moreover, D. gallinae is increasingly troubling layer farms in Europe (Sparagano et al 2009), at least partly due to important restrictions in the Maximal Residue Limits in the eggs (less acaricide products are allowed to be used during layer flocks, which are around one year long). On the other hand, Tucci et al (2008) suggest that D. gallinae is in the process of largely colonizing layer farms in Brazil. Ornithonyssus bursa was predominant in Brazilian hen farms around 1938-39 and this tendency was established inverted in favor of D. gallinae in 1997 (Tucci 1997). Moreover, the recurrent presence of D. gallinae in various bird groups noted in France by Roy et al (2009b), along with its spread in human-shaped environments, highlights the remarkable adaptive flexibility in this pest species.
Conclusion The utility of the intron n in Tropomyosin has been evidenced for inter- and intraspecific explorations within Dermanyssus. The split into hirsutus group and gallinae group in Dermanyssus does not seem to be convenient anymore, as D. hirsutus is branching from within the gallinae group. Within D. gallinae, several successive founder events seem to have occurred all along the fowl farming history. One bottleneck, likely due to pesticides and other farming practices, seems to have occurred formerly. More recently, some important admixing between long isolated populations are conspicuous in layer farm isolates. Commercial exchanges play an
160 Publication V important role in the dissemination of populations of the Poultry Red Mite D. gallinae, at least within the layer industry in France. Exchanges of mites between domestic and wild birds seem to occur very rarely to date (almost never). The genetic variability of D. gallinae in wild habitats is significantly higher than in other species, seemingly a consequence of an ancient radiation and of multiple hybridizations. The generalist D. gallinae seems to possess intrinsic characters of an invasive species, in contrast to the other four species, and showed the ability to colonize a variety of hosts, in very various habitats, successfully.
Cophylogenetic analyses would be interesting in order to establish more precisely the patterns of host switching and coevolution within Dermanyssus. Field experiments and population genetic analysis involving a larger amount of "focused isolates", in collaboration with the layer industry, are needed in order to establish the precise carrier(s) for mite dispersal in the framework of commercial exchanges. Also, the range of distance on which D. gallinae is able to move by itself (without the help of any other organism or vehicle) remains to be clearly estimated.
Supplementary material Appendix 1. Sampling and EMBL information for the populations under test in the present study Appendix 2. Information about the matrix of encoded In/Del (Tropomyosin) Appendix 3. ABC analysis of case II. 30 clusters of the 180 branching tree histories when considering 5 populations cladograms Appendix 4. Prior distribution and Bayesian probabilities of different scenarios in the three studies using an ABC method
Acknowledgments We want to warmly thank S. Lubac (Institut Technique de l’Aviculture, Lyon, France), F. Boléat et coll. (Ecopôle du Forez, France), O. Kilpinen (Danish Institute of Agricultural Sciences, Lyngby, Denmark), MW Sabelis and I Lesna (IBED, University of Amsterdam, The Netherlands), Damien and Marion Buronfosse for having provided precious mite samples. Many thanks to E. Vila (Maison de l'Orient et de la Méditerranée, Lyon, France) for valuable information about the history of hen domestication. Finally, LR would like to offer her most sincere thanks for their technical participation to S. Bonnet and N. Guichard N. (LEGTA Saint-Genis-Laval, France), M. Rigaux (IUT A, Université Lyon1, France), S. Merlin (Lycée Jean-Baptiste de la Salle, Lyon, France), S. El Ouartiti (IUT GBGE, Université Jean Monnet, Saint-Etienne, France), G. Lallemand (Lycée des Mandailles, Châteauneuf de Galaure, France). This work was supported by the PEP (Pôle d’Experimentation et de Progrès) Avicole of the Région Rhône-Alpes (France).
Literature cited
Arnold, M. 2004. Natural hybridization and the evolution of domesticated, pest and disease organisms. Mol Ecol 13:997-1007. Bandelt, H.J., P. Forster, A. and Röhl.1999. Median-joining networks for inferring intraspecific phylogenies. Mol Biol Evol 16:37-48. Beaumont, M.A., W. Zhang, and D. J. Balding. 2002. Approximate Bayesian computation in population genetics. Genetics 162:2025-2035.
161 Publication V
Beaumont, M. A. 2008. Joint determination of topology, divergence time, and immigration in population trees. In: Simulations, Genetics, and Human Prehistory. Edited by Matsumura S, Forster P, Renfrew C. Cambridge: McDonald Institute for Archaeological Research. p. 135-154. Berlese, A. and E. Trouessart. Diagnoses d’acariens nouveaux ou peu connus. 1889. Bull Bib Sci Ouest 2e année, 2e partie 9:121-143 Birky C. W., P. Fuerst, and T. Maruyama. 1989. Organelle gene diversity under migration, mutation and drift: equilibrium expectations, approach to equilibrium, effects of heteroplasmic cells, and comparison to nuclear genes. Genetics 121:613-627. Brännström, S., D. A. Morrison, J. G. Mattsson, and J. Chirico. 2008. Genetic differences in internal transcribed spacer 1 between Dermanyssus gallinae from wild birds and domestic chickens. Med Vet Entomol 22:152-155. Clayton, D. H., and D. M. Tompkins. 1994. Ectoparasite virulence is linked to mode of transmission. Proc Biol Sci 256:211-217. Clayton D. H., and D. M. Tompkins. 1995. Comparative effects of mites and lice on the reproductive success of Rock Doves (Columba livia). Parasitology 110:195-206. Combes C. 2000. Parasites, hosts, questions. In: Poulin R, Morand S, Skorping A, editors. Evolutionary Biology of Host-Parasite Relationships: Theory Meets Reality. Elsevier Science B.V. p. 1-8. Desdevises Y., S. Morand, and P. Legendre. 2002. Evolution and determinants of host specificity in the genus Lamellodiscus (Monogenea). Zool J of the Linnean Soc 77:431-443. De Geer, C. 1778. Mémoires pour servir à l'histoire des insectes. VII:111-112. Edgar, R. C. 2004. MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Res 32:1792-1797. Edmands S., H. V. Feaman, J. S. Harrison, and C. C. Tommermann. 2005. Genetic consequences of many generations of hybridization between divergent Copepod populations. J Hered 96:114-123. Edwards S. V., L. Liu, and D. K. Pearl. 2007. High-resolution species trees without concatenation. Proc Nat Am Soc 104:5936-5941. Estoup, A., M. A. Beaumont, F. Sennedot, C. Moritz, and J. M. Cornuet. 2004. Genetic analysis of complex demographic scenarios: spatially expanding populations of the cane toad, Bufo marinus. Evolution 58:2021-2036. Excoffier, L., G. Laval, and S. Schneider. 2005. Arlequin ver. 3.0: An integrated software package for population genetics data analysis. Evol Bioinform Online 1:47-50. Fagundes N., N. Ray, M. A. Beaumont, S. Neuenschwander, and F. M. Salzano. 2007. Statistical evaluation of alternative models of human evolution. Proc Nat Am Soc 104:17614-17619. Galtier, N., M. Gouy, and C. Gautier. 1996. SeaView and Phylo_win, two graphic tools for sequence alignment and molecular phylogeny. Comput Appl Biosci 12:543-548. Gorrochotegui-Escalante, N., M. L. De Lourdes Munoz, I. Fernandez-Salas, B. J. Beaty, and W. C. Black 4th. 2000. Genetic isolation by distance among Aedes aegypti populations along the northeastern coast of Mexico. Am J Trop Med Hyg 62:200-209. Hamilton, G., M. Currat, N. Ray, G. Heckel, M. Beaumont, and L. Excoffier. 2005. Bayesian estimation of recent migration rates after a spatial expansion. Genetics 170:409-417 Hare, M. P. 2001. Prospects for nuclear gene phylogeography. Trends Ecol Evol. 16:700-706. Hey, J., and R. Nielsen. 2007. Integration within the Felsenstein equation for improved Markov chain Monte Carlo methods in population genetics. Proc Nat Am Soc 104:2785-2790. Hickerson, M., E. Stahl, and H. A. Lessios. 2006. Test for simultaneous divergence using approximate bayesian computation. Evolution. 60:2435-2453. Hudson, R. R. 1990. Gene genealogies and the coalescent process. Oxf Surv Evol Biol 7:1-44. Hudson, R. R., M. Slatkin, and W. P. Maddison. 1992. Estimation of levels of gene flow from DNA sequence data. Genetics 132:583-589. Huelsenbeck, J.P., and F. Ronquist. 2001. MrBayes: Bayesian Inference for Phylogenetic Trees. Bioinformatics 17:754-755. Hutcheson, H. J., and J. H. Oliver Jr. 1988. Spermiogenesis and reproductive biology of Dermanyssus gallinae (De Geer) (Parasitiformes: Dermanyssidae). J Med Entomol 25:321-330. Ihaka, R., and R. Gentleman. 1996. A language for data analysis and graphics. J Comput. Graph. Stat 5:299-314.
162 Publication V
Johnson, L. S., and D. J. Albrecht. 1993. Effects of haematophagous ectoparasites on nestling House Wrens, Troglodytes aedon: who pays the cost of parasitism? Oikos 66:255-262. Kaci-Chaouch, T., O. Verneau, and Y. Desdevises. 2008. Host specificity is linked to intraspecific variability in the genus Lamellodiscus (Monogenea). Parasitology 1-8. Kimura, M. 1969. The number of heterozygous nucleotide sites maintained in a finite population due to steady flux of mutations. Genetics 61:893-903. Kimura, M., T. Ohta. 1978. Stepwise mutation model and distribution of allelic frequencies in a finite population. Proc Nat Am Soc 75:2868-2872. Kilpinen, O. 2001. Activation of the poultry red mite, Dermanyssus gallinae (Acari: Dermanyssidae), by increasing temperatures. Exp Appl Acarol 25:859-867. Kilpinen, O. 2005. How to obtain a bloodmeal without being eaten by a host: the case of poultry red mite, Dermanyssus gallinae. Physiol Entomol 30:232–240. Klompen, H., M. Lekveishvili, and W. C. Black. 2007. Phylogeny of Parasitiform mites (Acari) based on rRNA. Mol Phyl Evol 43:936-951. Kuris, A. M., and K. D. Lafferty. 2000. Parasite-host modelling meets reality: adaptive peaks and their ecological attributes. In: Poulin R., S. Morand, and A. Skorping, editors. Evolutionary Biology of Host-Parasite Relationships: Theory Meets Reality, Elsevier Science BV p. 9-26. Lee, C. E. 2002. Evolutionary genetics of invasive species. Trends Ecol Evol 17:386-391. Legras, J., D. Merdinoglu, J. M. Cornuet, and F. Karst. 2007. Bread, beer and wine: Saccharomyces cerevisiae diversity reflects human history. Mol ecol 16:2091-2102. Librado, P., and J. Rozas. 2009. DnaSP v5: A software for comprehensive analysis of DNA polymorphism data. Bioinformatics 25: 1451-1452. Liu, L., D. K. Pearl, R. T. Brumfield, and S. V. Edwards. 2008. Estimating species trees using multiple-allele DNA sequence data. Evolution 62:2080-2091. Liu, Y. P., G. S. Wu, Y. G. Yao, Y. W. Miao, G. Luikart, M. Baig, A. Beja-Pereira, Z. L. Ding, M. G. Palanichamy, and Y. P. Zhang. 2006. Multiple maternal origins of chickens: Out of the Asian Jungles ? Mol Phyl Evol 38:12-19. Loader, C. R. 1996. Local Likelihood Density Estimation. Annals of Statistics 24:1602-1618. Lunt, D. H., D. X. Zhang, J. M. Szymura, and G. M. Hewitt. 1996. The insect cytochrome oxidase I gene: evolutionary patterns and conserved primers for phylogenetic studies. Insect Mol Biol 5:153-165. Marjoram, P., J. Molitor, V. Plagnol, and S. Tavare. 2009. Markov chain Monte Carlo without likelihoods. Proc Nat Am Soc 100:15324-15328. Mallet, J. 2005. Hybridization as an invasion of the genome. Trends Ecol Evol 20:229-237. McCracken, K., and M. Sorenson. 2005. Is homoplasy or lineage sorting the source of incongruent mtdna and nuclear gene trees in the stiff-tailed ducks (Nomonyx-oxyura)? Syst Biol 54:35-55. Michaux, J. R., P. Chevret, M. G. Filippucci, and M. Macholan. 2002. Phylogeny of the genus Apodemus with a special emphasis on the subgenus Sylvaemus using the nuclear IRBP gene and two mitochondrial markers: cytochrome b and 12SrRNA. Mol Phyl Evol 23:123-136. Miller, N., A. Estoup, S. Toepfer, D. Bourguet, L. Lapchin, S. Derridj, K. S. Kim, P. Reynaud, L. Furlan, and T. Guillemaud. 2005. Multiple Transatlantic Introductions of the Western Corn Rootworm. Science 310, 992. Moore, W. S. 1995. Inferring phylogenies from the mt-DNA variation: mitochondrial-gene trees versus nuclear-gene trees. Evolution 49:718-726. Moss, W. W., and J. H. Camin. 1970. Nest parasitism, productivity, and clutch size in Purple Martins. Science 168:1000-1003. Mullens, B. A., N. C. Hinkle, L. J. Robinson, and C. E. Szijj. 2001. Dispersal of Northern Fowl Mites, Ornithonyssus sylviarum, Among Hens in an Experimental Poultry House. J appl Poult Res 10:60-64. Nature Midi-Pyrénées (association). 2001. Etude et prospection du Martinet Pâle Apus pallidus en Midi-Pyrénées. http://www.premiumwanadoo.com/naturemp/6Proteger/Etudes/RapportMartinetPale2.pdf. 2009. 06.3
163 Publication V
Navajas, M., Y. Le Conte, M. Solignac, S. Cros-Arteil, and J. M. Cornuet. 2002. The complete sequence of the mitochondrial genome of the honeybee ectoparasite mite Varroa destructor (Acari: Mesostigmata) Mol Biol Evol 19: 2313-2317. Nichols, S. 2000. Gene trees and species trees are not the same. Trends Ecol Evol 16:358-364. Nisbet, A. J., J. F. Huntley, A. McKellar, N. Sparks, and R. McDevitt. 2006. A house dust mite allergen homologue from poultry red mite Dermanyssus gallinae (De Geer). Parasite Immunol 28:401-405. Nordborg, M. 2003. Coalescent theory. In Handbook of Statistical Genetics. 2nd edition. Edited by Edited by Balding, D.J., M. Bishop, and C. Cannings. Chichester: Wiley p. 602-635. Nordenfors, H., J. Höglund, and A. Uggla. 1999. Effects of temperature and humidity on oviposition, molting, and longevity of Dermanyssus gallinae (Acari: Dermanyssidae). J Med Entomol 36:68- 72. Nylander, J. A. A. 2004. MrModeltest v2. Program distributed by the author. Evolutionary Biology Centre, Uppsala University. Oliver, J. H. Jr. 1966. Notes on Reproductive Behavior in the Dermanyssidae. J Med Entomol 3:29-35. Pacejka, A. J., and C. F. Thompson. 1996. Does removal of old nests from nestboxes by researchers affect mite populations in subsequent nests of House Wrens? J Field Ornithol 67:558-564. Pacejka, J., C. M. Gratton, and C. F. Thompson. 1998. Do potentially virulent mites affect house wren reproductive success ? – Troglodytes aedon. Ecology 5:79. Palumbi, S. R., F. Cipriano, and M. P. Hare. 2001. Predicting nuclear gene coalescence from mitochondrial data: The three-times rule. Evolution 55:859-868. Pascual, M., M. Chapuis, F. Mestres, J. Balanya, and R. Huey. 2007. Introduction history of Drosophila subobscura in the New World: a microsatellite-based survey using ABC methods. Mol ecol 16:3069-3083. Radovsky, F. J. 1994. The Evolution of Parasitism and the Distribution of Some Dermanyssoid Mites (Mesostigmata) on Vertebrate Hosts. In Mites Ecological and Evolutionary Analyses of Life- History Patterns, ed. MA Houck, USA Reynaud, M. C., C. M. Chauve, and F. Beugnet. 1997. Dermanyssus gallinae (De Geer, 1778): Reproduction expérimentale du cycle et essai de traitement par la moxidectine et l’ivermectine. Rev Med Vet 148:433-438. Rieseberg, L. H., O. Raymond, D. M. Rosenthal, Z. Lai, K. Livingstone, T. Nakazato, J. L. Durphy, A; E. Schwarzbach, L. A. Donovan, and C. Lexer. 2003. Major ecological transitions in wild sunflowers facilitated by hybridization. Science 301:1211-1216. Rogers, S. M., and L. Bernatchez. 2007. The genetic architecture of ecological speciation and the association with signatures of selection in natural lake whitefish (Coregonus sp. Salmonidae) species pairs. Mol Biol Evol 24:1423-1438. Ronquist, F., and J. P. Huelsenbeck. 2003. MRBAYES 3: Bayesian phylogenetic inference under mixed models. Bioinformatics. 19:1572-1574. Roy, L., A. P. G. Dowling, C. M. Chauve, and T. Buronfosse. 2009a. Delimiting species boundaries within Dermanyssus Dugès, 1834 (Acari:Dermanyssidae) using a total evidence approach. Mol Phyl Evol 50:446-470. Roy, L., A. P. G. Dowling, C. M. Chauve, I. Lesna, M. W. Sabelis, and T. Buronfosse. 2009b. Molecular phylogenetic assessment of host range in five Dermanyssus species. Exp Appl Acarol in press Rozas, J., and R. Rozas. 1995. DnaSP, DNA sequence polymorphism: an interactive program for estimating Population Genetics parameters from DNA sequence data. Comput Applic Biosci 11:621-625. Schwenk, K., N. Brede, and B. Streit. 2008. Extent, processes and evolutionary impact of interspecific hybridization in animals. Phil. Trans. R. Soc B 363:2805-2811. Schweitzer, J. A., G. D. Martinsen, and T. G. Whitham. 2002. Cottonwood hybrids gain fitness traits of both parents: a mechanism for their long-term persistence? Am J Bot 89:981-990. Seehausen, O. 2004. Hybridization and adaptive radiation. Trends Ecol Evol 19:198-207. Seehausen, O, G. Takimoto, D. Roy, and J. Jokela. 2007. Speciation reversal and biodiversity dynamics with hybridization in changing environments. Mol Ecol 17:30-44.
164 Publication V
Siegel, P. B., A. Haberfeld, T. K. Mukherjee, L. C. Stallard, H. L. Marks, N. B. Anthony, and E. A. Dunnington. 1992. Jungle fowl–domestic fowl relationships: a use of DNA fingerprinting. World’s Poult. Sci J 48:147–155. Sparagano, O., A. Pavliüeviü, T. Murano, A. Camarda, H. Sahibi, O. Kilpinen, M. Mul, R. van Emous, S. le Bouquin, K. Hoel, and M. Cafiero. 2009. Prevalence and key figures for the poultry red mite. Dermanyssus gallinae infections in poultry farm systems. Exp Appl Acarol 48:3-10. Stone, J., and J. Repka. 1998. Using a Nonrecursive Formula to Determine Cladogram Probabilities. Syst Biol 47:617–624. Storz, J. F., and M. A. Beaumont. 2002. Testing for genetic evidence of population expansion and contraction: an empirical analysis of microsatellite DNA variation using a hierarchical Bayesian model. Evolution 56:154-166. Tucci, E. C. 1997. A laboratory method for the rearing of Dermanyssus gallinae (De Geer, 1778) (Acari, Dermanyssidae). Arq Inst Biol São Paulo 64:1-4. Tucci, E. C., A. P. Prado, and R. P. Araújo. 2008. Development of Dermanyssus gallinae (Acari: Dermanyssidae) at different temperatures. Vet Parasitol 155:127-132. Wakeley, J. 1996. The variance of pairwise nucleotide differences in two populations with migration. Theor Popul Biol 49:39-57. Weisheit, A. S., and P. D. Creighton. 1989. Interference by House Sparrows in nesting activities of Barn Swallows. J Field Ornithol 60:323-328. West, B., and B. X. Zhou. 1988. Did Chickens Go North ? New evidence for domestication. J Archeaol Sci 15:515-533. Witalinski, W. 2000. Aclerogamasus stenocornis sp. n., a fossil mite from the Baltic amber (Acari: Gamasida: Parasitidae). Genus 11:619-626. Wood, H. P. 1917. The chicken mite: its life history and habits. United States Department of Agriculture, Washington, DC., Bull 553:1-14. Yamashita, H., S. Okamoto, Y. Maeda, and T. Hashiguchi. 1994. Genetic relationships among domestic and jungle fowls revealed by DNA fingerprinting analysis. Jpn Poult Sci 31:335–344. Yee, T., and C. Wild. 1996. Vector generalized additive models. J. R. Stat Soc, B. 58:481-493. Zeman, P. 1988. Surface skin lipids of birds - a proper host kairomone and feeding inducer in the poultry red mite, Dermanyssus gallinae. Exp Appl Acarol 5:163-173.
165 Publication V
Tables
Primer name Sequence Gene portion SKPO-F 5' CTTTTTAGATCTTTAATTGAAA 3' COI RQ-COI-R 5' CCAGTAATACCTCCAATTGTAAAT 3' COI T5bis-F 5' TCGAGCACAGGAACATCACTG 3' Tropomyosin T5bis-R 5' AGTCTCGGCACGGTCTTCA 3' Tropomyosin
Table 1. Primer sequences.
Mt Mt Coalescence Nuclear Nuclear Nuclear external internal ratio monophyly monophyly monophyly branch branch support support support length length (Bayesian (Bootstrap (Bootstrap / Posterior / Bremer Bremer Probabilities) values on values on MP MP topologies) topologies) gap=5th state gap=missing D. gallinae 0.95 60/100 60/100 0.09 0.08 1 D. apodis 0.121 0.00 36 1.00 100/100 100/100 D. carpathicus 0.116 0.02 5 1.00 100/100 100/100 D. hirundinis 0.043 0.02 2 0.59 99/100 nm D. longipes 0.043 0.03 1 nm nm nm D. gallinae L1 0.073 0.01 8 0.97 96/92 96/100 D. gallinae except L1 0.026 0.06 0 nm nm nm D. gallinae Lmt1 0.9 0.01 90 nm nm nm D. gallinae Lmt2 0.01 0.03571 0.3 nm nm nm D. gallinae Lmt3 0.01 0.03125 0.3 nm nm nm D. gallinae+D. apodis nm nm - 0.91 100/93 100/100 D. hirundinis+D. nm nm nm carpathicus+D. longipes+D. hirsutus 0.175 0.42 0.4 D. 0.88 100/100 100/100 gallinae+D.apodis+D. hirsutus (clade D) nm nm -
Table 2. Evaluation of the three-times rule for species and other entities of Dermanyssus under test. nm = not monophyletic
166 Publication V
Tropomyosin mt-Co1 Nm (from Fst P value Fst) Fst P value Nm (from Fst) D. gallinae All Wild vs D. apodis 0.73835 <0.0001 0.17631 0.82812 <0.0001 0.10378 D. gallinae All Wild vs D. hirundinis DhirF 0.81934 <0.0001 0.12074 0.80102 <0.0001 0.12421 D. gallinae All Wild vs D. carpathicus 0.84537 <0.0001 0.09192 0.78628 <0.0001 0.13591 D. carpathicus vs D. apodis GO 0.96972 <0.0001 0.01561 0.92393 <0.0001 0.04117 D. hirundinis DhirF vs D. apodis GO 0.97856 <0.0001 0.01096 0.96457 <0.0001 0.01836 8020 vs 8006B5 0.39878 <0.0001 0.75381 8020 vs 8006B1 0.09610 0.0039 4.7027 0.83984 <0.0001 0.09535 8020 vs SK -0.00335 0.41699 f 0.91934 <0.0001 0.04387 8020 vs IL 0.19807 <0.0001 2.0243 0.89575 <0.0001 0.05819 SK vs 8006B5 0.58486 <0.0001 0.35490 SK vs 8006B1 0.10665 0.00098 4.1882 0.76838 <0.0001 0.15072 IL vs 8006B5 0.52019 <0.0001 0.46119 IL vs 8006B1 0.32236 <0.0001 1.0510 0.69884 <0.0001 0.21547 SK vs IL 0.25762 <0.0001 1.4409 0.87909 <0.0001 0.06877 9001 vs IL 0.43939 <0.0001 0.63793 0.97479 <0.0001 0.01293 9001 vs SK 0.65114 <0.0001 0.26789 0.98236 <0.0001 0.00898 9001 vs 8020 0.58870 <0.0001 0.34933 0.94154 <0.0001 0.03104 9001 vs 8006B1 0.65452 <0.0001 0.26392 0.95529 <0.0001 0.02340 D. gallinae L1 9001 vs D. apodis GO 0.96793 <0.0001 0.01657 0.98604 <0.0001 0.00708
Table 3. Pairwise Fst estimates between D. gallinae focused isolates and between specific datasets, corresponding P values and associated Nm for both nuclear and mitochondrial loci. For Tropomyosin, permutation tests of Fst estimates used the genotype as the randomization unit instead of the allele.
167 Publication V
Tropomyosin mt-Co1 (obs). S(obs). Number of P(h<=h obs) h (expected) h (expected) P(h <=h obs) Hd (expected) fifth state)(obs). P(Hd<=Hd obs) P(Hd <=Hd obs) Number of sequences. Number of sequences. Number of haplotypes, h. Number of segregating sites, Haplotype diversity, Hd (obs). Number of haplotypes, h (obs) Number of segregating sites, S. polymorphic/indel/missing sites, S Average number of differences, K. Number of haplotypes, h (gap as a Haplotype diversity, Hd (obs). Hd (expected) Average number of differences, K (obs). Average number of differences, K (gap as a fifth state)(obs). Isolate/population D. gallinae 8006B1 44 17 49 5 14.56 <0.0001 5 0.69767 0.8774 0.020 5.12051 14.73573 24 33 3 12.289 <0.0001 0.16304 0.90748 <0.0001 3.59058 D. gallinae 8020 40 18 50 5 14.58 <0.0001 5 0.65641 0.88805 0.003 5.52179 16.94231 19 21 3 9.347 0.002 0.44444 0.86594 0.001 4.25731 D. gallinae SK 44 14 41 3 13.08 <0.0001 3 0.54017 0.85677 0.002 4.60042 14.46195 24 3 4 3.57 1.00 0.30797 0.51839 0.107 0.32609 D. gallinae IL 40 37 116 18 19.63 0.329 18 0.94103 0.93624 0.488 9.56154 31.88077 20 4 5 4.13 1.00 0.74211 0.58455 0.895 1.05789 D. gallinae L1 9001 48 0 0 1 - - 1 0.00000 - - 0.00000 0.00000 24 8 2 6.284 0.004 0.22826 0.72339 0.003 1.82609 D. gallinae 8006B5 16 29 4 9.685 0.004 0.775 0.90519 0.374 14.725 D. apodis GO 40 4 9 2 5.20 <0.0001 2 0.50128 0.58449 0.268 2.00513 4.51154 20 3 4 3.524 1.00 0.28421 0.52435 0.081 0.38947 D. hirundinis DhirF 34 0 0 1 - - 1 0.00000 - - 0.00000 0.00000 21 15 4 8.335 0.020 0.53333 0.82551 0.010 4.59048 D. hirundinis FR+USA 38 8 10 3 8.74 <0.0001 3,00 0.19772 0.77086 <0.0001 1.40683 1.79374 19 29 3 10.681 <0.0001 0.50292 0.89884 0.001 8.94737 D. longipes 16 21 56 5 8.75 0.041 5,0 0.73333 0.879 0.037 10.46667 28.8 8 25 3 6.175 0.017 0.71429 0.91721 0.013 14.14286 D. carpathicus 26 11 16 10 9.819 0.657 11,0 0.86462 0.84488 0.567 3.44308 5.17231 16 36 9 10.403 0.299 0.81667 0.92184 0.044 13.84167 D. gallinae All Farms 194 28 60 14 27.667 <0.0001 14,0 0.70124 0.89674 0.007 5.84573 16.71428 122 84 22 35.587 0.007 0.83796 0.94216 0.003 18.79461 D. gallinae All Wild 60 56 137 27 27.163 0.546 29,0 0.95028 0.94833 0.437 10.8209 30.69605 33 85 14 20.538 0.027 0.90909 0.95553 0.047 23.45265
Table 4. Computer simulations of coalescent process (DnaSP v5) given the number of segregating sites S, assuming an intermediate level of recombination R=10 for Tropomyosin amplicon and no recombination for COI amplicon (confidence interval =95%).
168 Publication V
Parameters Description Mode 0.95 Ne1 Population size in French layer farms 2000 0 8000 Population size in non-French layer Ne2 farms 24000 0 42000 Population size in French non-hen layer Ne3 farms 79000 0 200000 Ne4 Population size in Wild birds 453000 82000 500000 NeA1 First ancestral population size 217000 0 337000 NeA2 Second ancestral population size 264000 29000 500000 NeA3 Third ancestral population size 43000 0 68000 m1 Migration in French layer farms 0 0 0.0094 m2 Migration in non-French layer farms 0.0032 0 0.0056 Migration in French non-hen layer m3 farms 0 0 0.0057 m4 Migration in Wild Birds 0.0002 0 0.0005 mA1 Migration in first ancestral population 0.0100 0.0006 0.0100 Migration in second ancestral mA2 population 0.0026 0 0.0043 t1 First splitting time 0 0 280000 t2 Second splitting time 590000 70000 1110000 t3 Third splitting time 1270000 470000 1590000
Table 5. Estimates of modes and 95% credible intervals for the considered demographic parameter for the D. gallinae groups with different host types (case I, popABC).
Paramete Description Mode 0.95 rs Ne1 8006 population size 0 0 421000 Ne2 8020 population size 0 0 33000 Ne3 IL population size 0 0 464000 Ne4 JBO population size 500000 277000 500000 Ne5 SK population size 0 0 31000 NeA1 First ancestral population size 500000 38000 500000 NeA2 Second ancestral population size 168000 0 248000 NeA3 Third ancestral population size 0 0 474000 NeA4 Fourth ancestral population size 122000 0 192000 m1 Migration in 8006 0 0 0.0022 m2 Migration in 8020 0.0020 0 0.0035 m3 Migration in IL 0 0 0.0016 m4 Migration in JBO 0.0070 0.0006 0.0097 m5 Migration in SK 0.0014 0 0.0025 mA1 Migration in first ancestral population 0.0019 0 0.0047 mA2 Migration in second ancestral population 0.0040 0 0.0069 mA3 Migration in third ancestral population 0.0032 0 0.0055 t1 First splitting time 0 0 860000 t2 Second splitting time 1020000 130000 2390000 t3 Third splitting time 2490000 750000 2500000 t4 Fourth splitting time 2500000 1130000 2500000
169 Publication V
Table 6. Estimates of modes and 95% credible intervals for the considered demographic parameter for the D. gallinae populations in different geographical locations (case II, popABC).
Paramete Discription Mode 0.95 rs Ne1 D. apodis population size 34000 0 93000 Ne2 D. gallinae population size 93000 67000 106000 Ne3 D. hirundinis population size 31000 0 95000 NeA1 First ancestral population size 68000 0 98000 NeA2 Second ancestral population size 68000 0 100000 t1 First splitting time 0 0 5070000 t2 Second splitting time 3680000 420000 8690000
Table 7. Estimates of modes and 95% credible intervals for the considered demographic parameter for the Dermanyssus species (case III, popABC).
n na G H(obs) HWE P(SE) SK 44 3 4 0.41 0.11555 (0.00108) 8020 40 4 7 0.65 0.41775 (0.00149) 8006 44 5 8 0.86 0.02825 (0.00052) IL 36 18 18 0.88 0.13373 (0.00041) GO 40 2 3 0.14 0.00374 (0.00018) DhirF 36 1 1 - - 9001 48 1 1 - -
Table 8. Genotypic and heterozygosity variability in focused isolates for Tropomyosin exon n, intron n and exon n+1. n refers to the number of sequences under test, na to the allele number, G to the genotype number, H(obs) to the observed percentage of heterozygozity.
170 Publication V
Figures Legends
Fig. 1. Overview of the distribution of variable elements along the studied Tropomyosin sequence (mutation points and insertion/deletion sites).
Fig. 2. Haplotypes distribution according to the six ecological categories as defined in § Material and Methods. Network 4.5.1.0, epsilon=10, post-processioning MP processed (shortest trees), 187 individuals of D. gallinae sequenced in both COI and Tropomyosin and additional individuals sequenced only in one of these loci (total: 228 individuals for COI and 203 for Tropomyosin). A. Haplotype network for COI. H_n reads Co_n in the text. B. Haplotype network for Tropomyosin. H_n reads Tro_n in the text. The size of circle is proportional to the number of individuals sharing this haplotype for COI and to the number of alleles representing this haplotype in Tropomyosin. The small squares correspond to missing intermediate haplotypes. The length of lines is proportional to the number of mutated positions.
Fig. 3. Amount of “pure synapomorphies” (i.e. synapomorphies strictly characterizing observed entity as opposed to all other) found in diverse entities of various taxonomic levels on targeted Tropomyosin sequences, based on Roy et al. (2009a)'s species boundaries and the individuals under test in present study.
Fig. 4. Haplotypic topologies obtained with COI and Tropomyosin sequences. A, B, C and D. Tromopyosin based topologies. A, B and C. Maximum parsimony criterion, PAUP4.0. A. Gaps treated as a fifth state. Strict consensus of 264 most parsimonious trees (L=1288 CI=0.8238 RI=0.9320). Dots indicate alleles which have been isolated from individuals which group within the three mt lineages tested in Table 2: grey triangles, Lmt1; white squares, Lmt2; black circles, Lmt3. Gaps treated as missing data. Strict consensus of 1000 most parsimonious trees (L=607 CI=0.8023 RI=0.9263). C. Gaps alone, encoded following appendix. Strict consensus of 1000 most parsimonious trees (L=80 CI=0.7250 RI=0.9450). D. Bayesian analysis with gaps as missing data…E and F. COI based topologies. E. Maximum parsimony criterion, PAUP4.0. Strict consensus of the 434 most parsimonious trees (L=749 CI=0.5340 RI=0.8452). F. Bayesian analysis
Figure 5. Most supported topologies for the population genetics analysis using ABC methods. A. D. gallinae groups with different host types (case I). B D. gallinae populations with different geographic locations (case II). C. Dermanyssus species (case III).
171 Publication V
172 Publication V
Appendix 1. Sampling and EMBL information for the populations under test in present study.
Both tropomyosin accession numbers are given in the two right columns in case of heterozygosity, in the first one in case of homozygosity (HOM. in the last column in such a case).
COI Tropomyosin haplotypes Dermanyssus Departement COI accession Tropomyosin accession species Host Context Country (France) Isolate Individual haplotype numbers numbers A. casalis France ACA ACA Co_Acasalis AM921868 on bird -bird D. apodis Swift care activity France 69 MAR mar 1 Co_52 AM921880 Tro_29 + Tro_29 FM897372 HOM. on bird -bird id. D. apodis Swift care activity France 69 MAR mar 2 Co_52 AM921880 Tro_29 + Tro_29 FN257768 HOM. bird nest - inside a building, in a id. id. D. apodis Swift town France 30 GO102 GO102a Co_53 xxxxxxxx Tro_28 + Tro_29 FM897373 FN257763 bird nest - inside a building, in a id. D. apodis Swift town France 30 GO108 GO108c Tro_28 + Tro_28 FM897373 HOM. bird nest - inside a building, in a id. D. apodis Swift town France 30 GO111 GO111 Tro_28 + Tro_28 FM897373 HOM. bird nest - inside a building, in a id. D. apodis Swift town France 30 GO40 GO40b Tro_28 + Tro_28 FM897373 HOM. bird nest - inside a building, in a id. id. D. apodis Swift town France 30 GO54 GO542 Co_54 AM921874 Tro_28 + Tro_29 FM897373 FN257763
173 Publication V
bird nest - inside a building, in a id. id. id. D. apodis Swift town France 30 GO54 GO543 Co_54 AM921874 Tro_28 + Tro_28 FM897373 FM897373 bird nest - inside a building, in a id. id. id. D. apodis Swift town France 30 GO54 GO54b Co_54 AM921874 Tro_28 + Tro_28 FM897373 FM897373 bird nest - inside a building, in a id. id. id. D. apodis Swift town France 30 GO54 GO54c Co_54 AM921874 Tro_28 + Tro_28 FM897373 FM897373 bird nest - inside a building, in a id. D. apodis Swift town France 30 GO58 GO58a Co_54 FM179370 Tro_29 + Tro_29 FN257763 HOM. bird nest - inside a building, in a D. apodis Swift town France 30 GO59 GO591 Co_54 xxxxxxxx Tro_29 + Tro_29 FN257763 FN257764 bird nest - inside a building, in a id. id. D. apodis Swift town France 30 GO59 GO592 Co_54 xxxxxxxx Tro_28 + Tro_29 FM897373 FN257763 bird nest - inside a building, in a D. apodis Swift town France 30 GO59 GO593 Co_54 xxxxxxxx Tro_28 + Tro_28 FM897373 HOM. bird nest - inside a building, in a D. apodis Swift town France 30 GO63 GO63a Co_54 xxxxxxxx + bird nest - inside a building, in a D. apodis Swift town France 30 GO64 GO64a Co_54 xxxxxxxx + bird nest - inside a id. id. D. apodis Swift building, in a France 30 GO69 GO69a Co_54 xxxxxxxx Tro_29 + Tro_29 FN257763 FN257763
174 Publication V
town
bird nest - inside a building, in a id. id. D. apodis Swift town France 30 GO69 GO69b Co_54 xxxxxxxx Tro_29 + Tro_29 FN257763 FN257763 bird nest - inside a building, in a id. D. apodis Swift town France 30 GO78 GO78a Co_54 xxxxxxxx Tro_28 + Tro_28 FM897373 HOM. bird nest - inside a building, in a id. D. apodis Swift town France 30 GO78 GO78b Co_52 xxxxxxxx Tro_28 + Tro_28 FM897373 HOM. bird nest - inside a building, in a id. id. D. apodis Swift town France 30 GO96 GO96d Co_54 xxxxxxxx Tro_29 + Tro_29 FN257763 FN257763 bird nest - inside a building, in a id. id. D. apodis Swift town France 30 GO96 GO96f Co_54 xxxxxxxx Tro_29 + Tro_29 FN257763 FN257763 bird nest - inside a building, in a D. apodis Swift town France 30 GO96 GO96h Co_54 xxxxxxxx bird nest - inside a building, in a id. id. D. apodis Swift town France 30 GO96 GO96i Co_54 xxxxxxxx Tro_29 + Tro_29 FN257763 FN257763 bird nest - inside a building, in a id. D. apodis Swift town France 30 GOCA GOCANB Co_55 xxxxxxxx Tro_28 + Tro_28 FM897373 HOM. D. bird nest - nest carpathicus Tit box France 69 Ecop1 Ecop12 Co_51 FM208731 Tro_37 + Tro_37 FN257833 HOM. D. bird nest - nest carpathicus Tit box France 69 Ecop3 Ecop3a Co_50 FM208729 Tro_38 + Tro_39 FN257827 FN257828 D. Tit bird nest - nest France 13 JBO133 JBO133f Co_45 xxxxxxxx Tro_35 Tro_35 id. HOM.
175 Publication V
carpathicus box +FN257829 D. bird nest - nest carpathicus Tit box France 13 JBO134 JBO134d Co_45 xxxxxxxx Tro_34 +Tro_35 FN257829 D. bird nest - nest id. carpathicus Tit box France 13 JBO134 JBO134f Co_45 xxxxxxxx Tro_35 +Tro_35 FN257829 HOM. D. bird nest - nest carpathicus Tit box France 13 JBO135 JBO135a Tro_34 + Tro_34 D. bird nest - nest carpathicus Tit box France 13 JBO135 JBO135B Co_45 xxxxxxxx Tro_36 +Tro_37 FN257830 FN257833 D. bird nest - nest id. carpathicus Tit box France 13 JBO135 JBO135f Co_45 xxxxxxxx Tro_35 +Tro_35 FN257829 HOM. D. bird nest - nest carpathicus Tit box JBO83 JBO83b Co_45 xxxxxxxx Tro_40 + Tro_40 FN257765 FN257832 D. carpathicus Tit layer farm France 62 JMCO10 JMC10A Co_49 AM943021 Tro_32 + Tro_32 FN257834 HOM. bird nest - D. within a hole in carpathicus Tit a house's wall France 42 Mes Mes1 Co_43 FM208715 Tro_32 + Tro_32 FN257831 HOM. bird nest - D. within a hole in carpathicus Tit a house's wall France 42 Mes Mes3 Co_48 FM208723 bird nest - D. girder within a carpathicus Redstart building France 42 RQ RQ18 Co_43 xxxxxxxx Tro_32 + Tro_33 FM897375 FN257836 bird nest - D. girder within a carpathicus Redstart building France 42 RQ RQ20 Co_44 xxxxxxxx Tro_30 + Tro_31 bird nest - D. girder within a carpathicus Redstart building France 42 RQ RQ22 Co_45 xxxxxxxx bird nest - D. girder within a carpathicus Redstart building France 42 RQ RQ23 Co_46 xxxxxxxx Tro_32 + Tro_33 bird nest - D. girder within a carpathicus Redstart building France 42 RQ RQ24 Co_47 xxxxxxxx layer farm - D. gallinae Layer hen cages France 26 8002 8002b Co_1 FM208713 Tro_3 + Tro_3 FN257814 HOM. D. gallinae Layer hen layer farm - France 38 8003 8003b1a Co_1 FM208733 Tro_1 Tro_1 FN257811 HOM.
176 Publication V
cages + layer farm - id. D. gallinae Layer hen cages France 38 8003 8003b1c Co_1 FM208733 Tro_3 + Tro_3 FN257812 HOM. amateur layer FM208722 D. gallinae Layer hen house France 42 8004 8004a Co_8 Tro_2 + Tro_2 FN257770 HOM. amateur layer id. D. gallinae Layer hen house France 42 8004 8004b Co_8 FM208722 Tro_2 + Tro_2 FM897374 HOM. amateur layer D. gallinae Layer hen house France 42 8004 8004e Tro_2 + Tro_2 FN257771 HOM. layer farm - on D. gallinae Layer hen ground France 1 8006B5 8006b Co_4 xxxxxxxx layer farm - on id. D. gallinae Layer hen ground France 1 8006B1 8006B16a Co_4 xxxxxxxx Tro_2 + Tro_2 FN257781 HOM. layer farm - on id. id. D. gallinae Layer hen ground France 1 8006B1 8006B16b Co_4 xxxxxxxx Tro_2 + Tro_16 FN257781 FN257813 layer farm - on id. id. D. gallinae Layer hen ground France 1 8006B1 8006B16c Co_4 xxxxxxxx Tro_2 + Tro_16 FN257781 FN257813 layer farm - on id. id. D. gallinae Layer hen ground France 1 8006B1 8006B16d Co_4 xxxxxxxx Tro_2 + Tro_16 FN257781 FN257813 layer farm - on id. id. D. gallinae Layer hen ground France 1 8006B1 8006B16e Co_4 xxxxxxxx Tro_2 + Tro_3 FN257781 FN257812 layer farm - on id. id. D. gallinae Layer hen ground France 1 8006B1 8006B16f Co_4 xxxxxxxx Tro_3 + Tro_2 FN257812 FN257781 layer farm - on id. D. gallinae Layer hen ground France 1 8006B1 8006B16g Co_4 xxxxxxxx Tro_2 + Tro_17 FN257781 xxxxxxxx layer farm - on id. id. D. gallinae Layer hen ground France 1 8006B1 8006B16h Co_4 xxxxxxxx Tro_3 + Tro_16 FN257812 FN257813 layer farm - on D. gallinae Layer hen ground France 1 8006B1 8006B16i Co_4 xxxxxxxx layer farm - on id. id. D. gallinae Layer hen ground France 1 8006B1 8006B16j Co_4 xxxxxxxx Tro_2 + Tro_16 FN257781 FN257813 layer farm - on id. id. D. gallinae Layer hen ground France 1 8006B1 8006B16k Co_4 xxxxxxxx Tro_3 + Tro_2 FN257812 FN257781 layer farm - on id. id. D. gallinae Layer hen ground France 1 8006B1 8006B16l Co_4 xxxxxxxx Tro_1 + Tro_2 FN257872 FN257781
177 Publication V
layer farm - on id. id. D. gallinae Layer hen ground France 1 8006B1 8006B1M2a Co_4 xxxxxxxx Tro_3 + Tro_1 FN257812 FN257872 layer farm - on D. gallinae Layer hen ground France 1 8006B1 8006B1M2b Co_4 xxxxxxxx layer farm - on id. D. gallinae Layer hen ground France 1 8006B1 8006B1M2c Co_1 xxxxxxxx Tro_17 + Tro_1 xxxxxxxx FN257872 layer farm - on id. id. D. gallinae Layer hen ground France 1 8006B1 8006B1M2d Co_4 xxxxxxxx Tro_2 + Tro_3 FN257781 FN257812 layer farm - on id. id. D. gallinae Layer hen ground France 1 8006B1 8006B1M2e Co_4 xxxxxxxx Tro_16 + Tro_2 FN257813 FN257781 layer farm - on id. D. gallinae Layer hen ground France 1 8006B1 8006B1M2f Co_4 xxxxxxxx Tro_17 + Tro_2 xxxxxxxx FN257781 layer farm - on id. id. D. gallinae Layer hen ground France 1 8006B1 8006B1M2g Co_4 xxxxxxxx Tro_2 + Tro_1 FN257781 FN257872 layer farm - on id. D. gallinae Layer hen ground France 1 8006B1 8006B1M2h Co_4 xxxxxxxx Tro_17 + Tro_2 xxxxxxxx FN257781 layer farm - on id. id. D. gallinae Layer hen ground France 1 8006B1 8006B1M2i Co_4 xxxxxxxx Tro_2 + Tro_3 FN257781 FN257812 layer farm - on id. id. D. gallinae Layer hen ground France 1 8006B1 8006B1M2j Co_4 xxxxxxxx Tro_2 + Tro_3 FN257781 FN257812 layer farm - on id. id. D. gallinae Layer hen ground France 1 8006B1 8006B1M2kCo_2 xxxxxxxx Tro_2 + Tro_2 FN257781 FN257781 layer farm - on id. id. D. gallinae Layer hen ground France 1 8006B1 8006B1M2l Co_4 xxxxxxxx Tro_2 + Tro_2 FN257781 FN257781 layer farm - on D. gallinae Layer hen ground France 1 8006B5 8006d Co_4 xxxxxxxx layer farm - on D. gallinae Layer hen ground France 1 8006B5 8006e Co_1 xxxxxxxx layer farm - on D. gallinae Layer hen ground France 1 8006B5 8006f Co_3 xxxxxxxx layer farm - on D. gallinae Layer hen ground France 1 8006B5 8006g Co_1 xxxxxxxx layer farm - on D. gallinae Layer hen ground France 1 8006B5 8006i Co_4 xxxxxxxx
178 Publication V
layer farm - on D. gallinae Layer hen ground France 1 8006B5 8006j Co_5 xxxxxxxx Tro_2 + Tro_2 FN257773 HOM. layer farm - on D. gallinae Layer hen ground France 1 8006B5 8006k Co_3 xxxxxxxx layer farm - on D. gallinae Layer hen ground France 1 8006B5 8006l Co_4 xxxxxxxx layer farm - on D. gallinae Layer hen ground France 1 8006B5 8006m Co_4 xxxxxxxx layer farm - on D. gallinae Layer hen ground France 1 8006B5 8006n Co_4 xxxxxxxx layer farm - on D. gallinae Layer hen ground France 1 8006B5 8006p Co_5 xxxxxxxx layer farm - on D. gallinae Layer hen ground France 1 8006B5 8006q Co_3 xxxxxxxx layer farm - on D. gallinae Layer hen ground France 1 8006B5 8006r Co_1 xxxxxxxx layer farm - on D. gallinae Layer hen ground France 1 8006B5 8006s Co_1 xxxxxxxx layer farm - on D. gallinae Layer hen ground France 1 8006B5 8006t Co_5 xxxxxxxx
layer farm - on D. gallinae Layer hen ground Belgium 8007 8007a Co_6 FM208717 Tro_3 + Tro_21 FN257820 FN257821
layer farm - on id. D. gallinae Layer hen ground Belgium 8007 8007b Co_6 FM208717 Tro_3 + Tro_3 FN257815 HOM. pigeon breeding D. gallinae Pigeon facility - aviary 69 8008 8008g Co_34 FM208712 Tro_8 + Tro_10 FN257787 FN257788 pigeon breeding id. D. gallinae Pigeon facility - aviary 69 8008 8008h Co_34 FM208712 Tro_8 + Tro_10 FN257789 FN257790 pigeon breeding id. D. gallinae Pigeon facility - aviary 69 8008 8008i Co_34 FM208712 Tro_8 + Tro_10 FN257791 FN257792 D. gallinae Layer hen layer farm - France 69 8009 8009a Co_1 FM208724 Tro_20 Tro_2 FN257774 FN257775
179 Publication V
cages + layer farm - on D. gallinae Layer hen ground France 26 8010 8010a Co_1 xxxxxxxx Tro_16 + Tro_16 FN257813 HOM. layer farm - on D. gallinae Layer hen ground France 26 8010 8010c Co_1 xxxxxxxx Tro_2 + Tro_2 FN257776 HOM. layer farm - D. gallinae Layer hen cages France 26 8011 8011a Co_1 xxxxxxxx Tro_2 + Tro_2 FN257777 HOM. layer farm - D. gallinae Layer hen cages France 26 8011 8011b Co_1 xxxxxxxx Tro_2 + Tro_2 FN257778 HOM. layer farm - D. gallinae Layer hen cages France 26 8011 8011c Co_1 xxxxxxxx Tro_2 + Tro_2 FN257779 HOM. chicken farm - Bresse D. gallinae Layer hen (epinettes) France 1 8012 8012a Co_7 FM208739 Tro_3 + Tro_3 FN257822 HOM. chicken farm - Bresse id. D. gallinae Layer hen (epinettes) France 1 8012 8012d Co_7 FM208739 Tro_3 + Tro_3 FN257816 HOM. layer farm - D. gallinae Layer hen cages France 7 8018 8018a Co_1 xxxxxxxx Tro_19 +Tro_2 FN257780 FN257781 layer farm - D. gallinae Layer hen cages France 7 8018 8018b Co_1 xxxxxxxx Tro_2 + Tro_2 FN257782 HOM. layer farm - on D. gallinae Layer hen ground France 7 8019 8019a Co_1 xxxxxxxx Tro_2 + Tro_2 FN257783 HOM.
D. gallinae Layer hen layer farm France 49 8020 8020a Tro_2 + Tro_2 FN257867 HOM.
D. gallinae Layer hen layer farm France 49 8020 8020b Co_1 xxxxxxxx Tro_2 + Tro_1 FN257785 FN257786
D. gallinae Layer hen layer farm France 49 8020 8020c Co_1 xxxxxxxx Tro_2 + Tro_1 FN257868 FN257869 id. id. D. gallinae Layer hen layer farm France 49 8020 8020d Tro_2 + Tro_1 FN257781 FN257872
D. gallinae Layer hen layer farm France 49 8020 8020e Co_1 xxxxxxxx Tro_3 + Tro_3 FN257817 HOM.
D. gallinae Layer hen layer farm France 49 8020 8020f Co_9 xxxxxxxx Tro_16 + Tro_1 FN257870 FN257871 id. id. D. gallinae Layer hen layer farm France 49 8020 8020g Co_1 xxxxxxxx Tro_2 + Tro_1 FN257781 FN257872 D. gallinae Layer hen layer farm France 49 8020 8020h Co_1 xxxxxxxx Tro_2 Tro_2 id. HOM.
180 Publication V
+FN257781 id. id. D. gallinae Layer hen layer farm France 49 8020 8020i Co_1 xxxxxxxx Tro_2 + Tro_1 FN257781 FN257872 id. D. gallinae Layer hen layer farm France 49 8020 8020j Co_9 xxxxxxxx Tro_2 + Tro_2 FN257781 HOM. id. id. D. gallinae Layer hen layer farm France 49 8020 8020k Co_1 xxxxxxxx Tro_2 + Tro_1 FN257781 FN257872
D. gallinae Layer hen layer farm France 49 8020 8020l Co_1 xxxxxxxx Tro_1 +Tro_1 FN257872 HOM.
D. gallinae Layer hen layer farm France 49 8020 8020m Co_10 xxxxxxxx Tro_3 + Tro_2 FN257873 FN257874 id. D. gallinae Layer hen layer farm France 49 8020 8020n Co_1 xxxxxxxx Tro_2 + Tro_2 FN257781 HOM. id. id. D. gallinae Layer hen layer farm France 49 8020 8020o Co_10 xxxxxxxx Tro_2 + Tro_1 FN257781 FN257872 id. id. D. gallinae Layer hen layer farm France 49 8020 8020p Co_1 xxxxxxxx Tro_3 + Tro_1 FN257812 FN257872 id. id. D. gallinae Layer hen layer farm France 49 8020 8020q Co_1 xxxxxxxx Tro_2 + Tro_1 FN257781 FN257872
D. gallinae Layer hen layer farm France 49 8020 8020r Co_1 xxxxxxxx + id. D. gallinae Layer hen layer farm France 49 8020 8020s Co_1 xxxxxxxx Tro_1 + Tro_1 FN257872 HOM. id. id. D. gallinae Layer hen layer farm France 49 8020 8020t Co_1 xxxxxxxx Tro_3 + Tro_2 FN257812 FN257781 id. id. D. gallinae Layer hen layer farm France 49 8020 8020u Co_10 xxxxxxxx Tro_1 + Tro_2 FN257872 FN257781 id. id. D. gallinae Layer hen layer farm France 49 8022 8022a Tro_3 + Tro_2 FN257812 FN257781 pigeon breeding id. id. D. gallinae Pigeon facility - aviary 69 9001 9001a Co_35 AM921875 Tro_10 + Tro_10 FN257788 HOM. pigeon breeding id. id. D. gallinae Pigeon facility - aviary 69 9001 9001b Co_34 FM208712 Tro_10 + Tro_10 FN257788 HOM. pigeon breeding id. id. D. gallinae Pigeon facility - aviary 69 9001 9001c Co_34 FM208712 Tro_10 + Tro_10 FN257788 HOM.
181 Publication V
pigeon breeding id. id. D. gallinae Pigeon facility - aviary 69 9001 9001d Co_34 FM208712 Tro_10 + Tro_10 FN257788 HOM. pigeon breeding id. id. D. gallinae Pigeon facility - aviary 69 9001 9001e Co_34 FM208712 Tro_10 + Tro_10 FN257788 HOM. pigeon breeding id. id. D. gallinae Pigeon facility - aviary 69 9001 9001f Co_34 FM208712 Tro_10 + Tro_10 FN257788 HOM. pigeon breeding id. id. D. gallinae Pigeon facility - aviary 69 9001 9001g Co_34 FM208712 Tro_10 + Tro_10 FN257788 HOM. pigeon breeding id. id. D. gallinae Pigeon facility - aviary 69 9001 9001h Co_34 FM208712 Tro_10 + Tro_10 FN257788 HOM. pigeon breeding id. id. D. gallinae Pigeon facility - aviary 69 9001 9001i Co_34 FM208712 Tro_10 + Tro_10 FN257788 HOM. pigeon breeding id. id. D. gallinae Pigeon facility - aviary 69 9001 9001j Co_35 AM921875 Tro_10 + Tro_10 FN257788 HOM. pigeon breeding id. id. D. gallinae Pigeon facility - aviary 69 9001 9001k Co_34 FM208712 Tro_10 + Tro_10 FN257788 HOM. pigeon breeding id. id. D. gallinae Pigeon facility - aviary 69 9001 9001l Co_34 FM208712 Tro_10 + Tro_10 FN257788 HOM. pigeon breeding id. id. D. gallinae Pigeon facility - aviary 69 9001 9001m Co_34 FM208712 Tro_10 + Tro_10 FN257788 HOM. pigeon breeding id. id. D. gallinae Pigeon facility - aviary 69 9001 9001n Co_34 FM208712 Tro_10 + Tro_10 FN257788 HOM. pigeon breeding id. id. D. gallinae Pigeon facility - aviary 69 9001 9001o Co_34 FM208712 Tro_10 + Tro_10 FN257788 HOM. pigeon breeding id. id. D. gallinae Pigeon facility - aviary 69 9001 9001p Co_35 AM921875 Tro_10 + Tro_10 FN257788 HOM.
182 Publication V
pigeon breeding id. id. D. gallinae Pigeon facility - aviary 69 9001 9001q Co_34 FM208712 Tro_10 + Tro_10 FN257788 HOM. pigeon breeding id. id. D. gallinae Pigeon facility - aviary 69 9001 9001r Co_34 FM208712 Tro_10 + Tro_10 FN257788 HOM. pigeon breeding id. id. D. gallinae Pigeon facility - aviary 69 9001 9001s Co_34 FM208712 Tro_10 + Tro_10 FN257788 HOM. pigeon breeding id. id. D. gallinae Pigeon facility - aviary 69 9001 9001t Co_34 FM208712 Tro_10 + Tro_10 FN257788 HOM. pigeon breeding id. id. D. gallinae Pigeon facility - aviary 69 9001 9001u Co_34 FM208712 Tro_10 + Tro_10 FN257788 HOM. pigeon breeding id. id. D. gallinae Pigeon facility - aviary 69 9001 9001v Co_34 FM208712 Tro_10 + Tro_10 FN257788 HOM. pigeon breeding id. id. D. gallinae Pigeon facility - aviary 69 9001 9001w Co_34 FM208712 Tro_10 + Tro_10 FN257788 HOM. pigeon breeding id. id. D. gallinae Pigeon facility - aviary 69 9001 9001x Co_34 FM208712 Tro_10 + Tro_10 FN257788 HOM. chicken farm - Bresse D. gallinae Layer hen (epinettes) France 1 BER BER1 Co_15 xxxxxxxx Tro_1 + Tro_3 FN257877 FN257878 chicken farm - Bresse id. id. D. gallinae Layer hen (epinettes) France 1 BER BER2 Co_15 xxxxxxxx Tro_2 + Tro_3 FN257781 FN257812 chicken farm - Bresse id. id. D. gallinae Layer hen (epinettes) France 1 BER BER3 Co_15 xxxxxxxx Tro_1 + Tro_3 FN257872 FN257812 Various D. gallinae pet birds bird pet shop France 69 CANIM CANIM Co_34 FM208734 Tro_9 + Tro_8 FN257799 FN257800
D. gallinae Layer hen layer farm France 1 Chab Chab3 Co_16 xxxxxxxx Tro_2 + Tro_2 FN257806 HOM.
D. gallinae Layer hen layer farm France 1 Chab Chab4 Co_17 xxxxxxxx Tro_2 + Tro_2 FN257807 HOM.
183 Publication V
D. gallinae Layer hen layer farm France 1 Chab Chab5 Co_16 xxxxxxxx Tro_2 + Tro_2 FN257808 HOM. on bird - care D. gallinae Pigeon activity France 69 COL COL1 Co_35 AM921875 Tro_8 + Tro_8 FN257803 HOM. on bird - care id. D. gallinae Pigeon activity France 69 COL COL3 Co_35 AM921875 Tro_11 + Tro_12 FN257804 FN257805 layer farm - D. gallinae Layer hen cages France 29 F29 F29b Co_1 xxxxxxxx Tro_2 + Tro_2 FN257784 HOM. layer farm - D. gallinae Layer hen cages France 38 F38 F38_6002 Co_1 xxxxxxxx Tro_4 +Tro_1 FN257809 FN257810 layer farm - id. D. gallinae Layer hen cages France 56 F56 F56b Co_1 xxxxxxxx Tro_1 +Tro_1 FN257872 HOM. European bird nest - nest D. gallinae starling box Netherlands IL202 IL202A Co_27 xxxxxxxx Tro_27 + Tro_1 FN257902 FN257903 European bird nest - nest D. gallinae starling box Netherlands IL202 IL202C Co_28 xxxxxxxx Tro_23 + Tro_23 FN257826 HOM. European bird nest - nest D. gallinae starling box Netherlands IL202 IL202D Tro_23 + Tro_22 FN257883 FN257884 European bird nest - nest D. gallinae starling box Netherlands IL202 IL202m Co_28 xxxxxxxx European bird nest - nest D. gallinae starling box Netherlands IL202 IL202n Co_27 xxxxxxxx European bird nest - nest D. gallinae starling box Netherlands IL202 IL202o Co_28 xxxxxxxx Tro_1 + Tro_5 FN257885 FN257886 European bird nest - nest D. gallinae starling box Netherlands IL202 IL202p Co_28 xxxxxxxx Tro_51 +Tro_52 xxxxxxxx xxxxxxxx European bird nest - nest D. gallinae starling box Netherlands IL202 IL202q Co_26 xxxxxxxx Tro_53 + Tro_53 xxxxxxxx HOM. European bird nest - nest D. gallinae starling box Netherlands IL213 IL2131 Tro_49 + Tro_27 FN257910 FN257911 European bird nest - nest D. gallinae starling box Netherlands IL213 IL2132 Co_27 FM207499 Tro_50 + Tro_22 FN257881 FN257882 European bird nest - nest id. D. gallinae starling box Netherlands IL213 IL2133 Co_27 FM207499 Tro_50 + Tro_27 FN257906 FN257907 European bird nest - nest D. gallinae starling box Netherlands IL227 IL2272 Co_26 FM207496 Tro_27 + Tro_27 FN257895 HOM. European bird nest - nest id. D. gallinae starling box Netherlands IL227 IL2273 Co_26 FM207496 Tro_6 + Tro_7 FN257896 FN257897 D. gallinae European bird nest - nest Netherlands IL302 IL302A Co_27 xxxxxxxx
184 Publication V
starling box European bird nest - nest D. gallinae starling box Netherlands IL302 IL302B Co_29 xxxxxxxx European bird nest - nest D. gallinae starling box Netherlands IL302 IL302e Tro_5 + Tro_6 FN257887 FN257888 European bird nest - nest D. gallinae starling box Netherlands IL302 IL302f Co_30 xxxxxxxx Tro_1 + Tro_54 FN257908 FN257909 European bird nest - nest D. gallinae starling box Netherlands IL302 IL302h Co_26 xxxxxxxx Tro_49 +Tro_55 FN257912 FN257913 European bird nest - nest D. gallinae starling box Netherlands IL302 IL302m Co_26 xxxxxxxx Tro_6 +Tro_7 FN257889 FN257890 European bird nest - nest D. gallinae starling box Netherlands IL302 IL302n Co_26 xxxxxxxx Tro_27 + Tro_2 FN257900 FN257901 European bird nest - nest D. gallinae starling box Netherlands IL302 IL302o Co_27 xxxxxxxx Tro_7 + Tro_27 FN257891 FN257892 European bird nest - nest D. gallinae starling box Netherlands IL302 IL302p Co_27 xxxxxxxx Tro_6 + Tro_24 FN257893 FN257894 European bird nest - nest D. gallinae starling box Netherlands IL302 IL302q Co_27 xxxxxxxx Tro_56 + Tro_57 FN257898 FN257899 bird nest - nest D. gallinae Tit box France 13 JBO46 JBO461 Co_31 FM208736 Tro_26 + Tro_3 FN257823 FN257824 bird nest - nest id. D. gallinae Tit box France 13 JBO46 JBO464 Co_31 FM208736 Tro_26 + Tro_26 FN257825 HOM. bird nest - nest D. gallinae Tit box France 13 JBO75 JBO75a Co_32 xxxxxxxx bird nest - nest D. gallinae Tit box France 13 JBO75 JBO75b Co_32 xxxxxxxx Tro_3 + Tro_60 FN257904 FN257905 bird nest - nest D. gallinae Tit box France 13 JBO75 JBO75c Co_32 xxxxxxxx Tro_59 + Tro_3 FN257916 FN257917 bird nest - nest D. gallinae Tit box France 13 JBO90 JBO90a Co_32 xxxxxxxx bird nest - nest D. gallinae Tit box France 13 JBO90 JBO90b Co_32 xxxxxxxx bird nest - nest D. gallinae Tit box France 13 JBO90 JBO90c Co_32 xxxxxxxx House bird nest - from D. gallinae martin barn France 18 LB18 LB181 Co_21 xxxxxxxx Tro_1 +Tro_1 FN257801 HOM. House bird nest - from AM921867 D. gallinae martin barn France 18 LB18 LB183 Co_22 Tro_1 + Tro_1 FN257802 HOM.
185 Publication V
pigeon breeding D. gallinae Pigeon facility - aviary France 26 LC LC1 Co_36 AM921859 Tro_13 + Tro_11 FN257797 FN257798 pigeon breeding id. D. gallinae Pigeon facility - aviary France 26 LC LC4 Co_36 AM921859 Tro_11 + Tro_11 FN257793 HOM. bird nest - in D. gallinae Pigeon town France 13 PI Pl1 Co_35 AM921860 Tro_11 + Tro_11 FN257794 HOM. bird nest - in id. D. gallinae Pigeon town France 13 PI Pl2 Co_35 AM921860 Tro_14 + Tro_11 FN257795 FN257796
D. gallinae Layer hen layer farm Poland PO1 PO1b Co_18 xxxxxxxx Tro_3 +Tro_3 FN257818 HOM.
D. gallinae Layer hen layer farm Poland PO2 PO2A Co_18 xxxxxxxx Tro_3 + Tro_3 FN257819 HOM.
D. gallinae Layer hen layer farm Poland PO2 PO2B Co_19 xxxxxxxx Tro_2 + Tro_1 FN257875 FN257876 on bird - nest id. D. gallinae Roller box France 13 ROL1 ROL12 Tro_15 + Tro_2 xxxxxxxx FN257781 on bird - nest D. gallinae Roller box France 13 ROL1 ROL15 Co_23 AM921864 + on bird - nest id. D. gallinae Roller box France 13 ROL1 ROL16 Tro_15 + Tro_1 xxxxxxxx FN257872 on bird - nest D. gallinae Roller box France 13 ROL2 ROL23 Co_24 AM921865 on bird - nest id. D. gallinae Roller box France 13 ROL2 ROL25 Tro_15 + Tro_3 xxxxxxxx FN257812 on bird - nest D. gallinae Roller box France 13 ROL2 ROL26 Co_25 xxxxxxxx Tro_58 + Tro_15 FN257914 FN257915 amateur layer D. gallinae Layer hen house France 69 SB SB1 Co_20 AM921858 Tro_2 + Tro_2 FN257766 HOM. lab strain (original isolate from a layer D. gallinae Layer hen farm) Denmark SK SK_Fa17 Co_11 xxxxxxxx
186 Publication V
lab strain (original isolate from a layer D. gallinae Layer hen farm) Denmark SK SK_Fa18 Co_11 xxxxxxxx lab strain (original isolate from a layer D. gallinae Layer hen farm) Denmark SK SK081 Co_11 AM921856 Tro_3 + Tro_2 FN257861 FN257862 lab strain (original isolate from a layer id. id. id. D. gallinae Layer hen farm) Denmark SK SK0810 Co_11 AM921856 Tro_2 + Tro_3 FN257781 FN257812 lab strain (original isolate from a layer id. id. D. gallinae Layer hen farm) Denmark SK SK0811 Co_11 AM921856 Tro_2 + Tro_2 FN257781 HOM. lab strain (original isolate from a layer id. id. id. D. gallinae Layer hen farm) Denmark SK SK0812 Co_11 AM921856 Tro_2 + Tro_1 FN257781 FN257872 lab strain (original isolate from a layer id. D. gallinae Layer hen farm) Denmark SK SK0813 Co_11 AM921856 Tro_1 + Tro_1 FN257866 HOM. lab strain (original isolate from a layer id. id. D. gallinae Layer hen farm) Denmark SK SK0814 Co_11 AM921856 Tro_1 + Tro_1 FN257872 HOM. lab strain (original isolate from a layer id. D. gallinae Layer hen farm) Denmark SK SK0815 Co_12 xxxxxxxx Tro_2 + Tro_2 FN257781 HOM. lab strain (original isolate from a layer id. id. id. D. gallinae Layer hen farm) Denmark SK SK0816 Co_11 AM921856 Tro_2 + Tro_1 FN257781 FN257872 lab strain (original isolate id. id. D. gallinae Layer hen from a layer Denmark SK SK0817 Co_11 AM921856 Tro_2 + Tro_2 FN257781 HOM.
187 Publication V
farm)
lab strain (original isolate from a layer id. D. gallinae Layer hen farm) Denmark SK SK0818 Co_13 xxxxxxxx Tro_2 + Tro_2 FN257781 HOM. lab strain (original isolate from a layer id. D. gallinae Layer hen farm) Denmark SK SK0819 Co_14 xxxxxxxx Tro_2 + Tro_2 FN257781 HOM. lab strain (original isolate from a layer id. id. id. D. gallinae Layer hen farm) Denmark SK SK082 Co_11 AM921856 Tro_3 + Tro_2 FN257812 FN257781 lab strain (original isolate from a layer id. id. D. gallinae Layer hen farm) Denmark SK SK0820 Co_11 AM921856 Tro_2 + Tro_2 FN257781 HOM. lab strain (original isolate from a layer id. D. gallinae Layer hen farm) Denmark SK SK0821 Co_13 xxxxxxxx Tro_2 + Tro_2 FN257781 HOM. lab strain (original isolate from a layer id. id. D. gallinae Layer hen farm) Denmark SK SK0822 Co_11 AM921856 Tro_2 + Tro_2 FN257781 HOM. lab strain (original isolate from a layer id. D. gallinae Layer hen farm) Denmark SK SK0823 Tro_1 + Tro_1 FN257872 HOM. lab strain (original isolate from a layer id. id. D. gallinae Layer hen farm) Denmark SK SK0824 Co_11 AM921856 Tro_1 + Tro_1 FN257872 HOM. lab strain (original isolate from a layer id. D. gallinae Layer hen farm) Denmark SK SK0826 Co_11 AM921856
188 Publication V
lab strain (original isolate from a layer id. id. id. D. gallinae Layer hen farm) Denmark SK SK083 Co_11 AM921856 Tro_2 + Tro_1 FN257781 FN257872 lab strain (original isolate from a layer id. D. gallinae Layer hen farm) Denmark SK SK084 Co_11 AM921856 Tro_2 + Tro_2 FN257863 HOM. lab strain (original isolate from a layer id. D. gallinae Layer hen farm) Denmark SK SK085 Co_11 AM921856 lab strain (original isolate from a layer id. D. gallinae Layer hen farm) Denmark SK SK086 Co_11 AM921856 lab strain (original isolate from a layer id. D. gallinae Layer hen farm) Denmark SK SK087 Co_11 AM921856 Tro_2 + Tro_1 FN257864 FN257865 lab strain (original isolate from a layer id. id. id. D. gallinae Layer hen farm) Denmark SK SK088 Co_11 AM921856 Tro_2 + Tro_1 FN257781 FN257872 lab strain (original isolate from a layer id. id. id. D. gallinae Layer hen farm) Denmark SK SK089 Co_11 AM921856 Tro_3 + Tro_2 FN257812 FN257781 on bird - banding D. gallinae Pic activity France 69 Woodp Woodp Co_33 AM921863 Tro_25 + Tro_25 xxxxxxxx HOM.
D. hirsutus USA ADhirs ADhirs Co_Dhirsutus AM921878 Tro_Dhirsutus + Tro_Dhirsutus FM897371 HOM. Barn bird nest - from D. hirundinis swallow barn France 1 CB4 CB4d Co_38 FM208727 Tro_44 + Tro_44 FN257838 HOM. Barn bird nest - from D. hirundinis swallow barn France 1 CB5 CB5c Co_38 FM208728 Tro_44 + Tro_44 FN257839 HOM. Barn bird nest - from id. D. hirundinis swallow barn France 1 CB5 CB5d Co_38 FM208728 Tro_44 + Tro_44 FN257850 HOM.
189 Publication V
Barn bird nest - from id. D. hirundinis swallow barn France 1 CB5 CB5e Co_38 FM208728 Tro_44 + Tro_44 FN257840 HOM. Barn bird nest - from D. hirundinis swallow barn France 72 CHOV CHOV1 Co_38 FM179369 Tro_44 + Tro_44 FN257841 HOM. Barn bird nest - from id. D. hirundinis swallow barn France 72 CHOV CHOV2 Co_38 FM179369 Tro_44 + Tro_44 FN257842 HOM. Barn bird nest - from D. hirundinis swallow barn France 85 HIR1 HIR1A Co_39 FM179366 Tro_44 + Tro_44 FN257843 HOM. Barn bird nest - from id. D. hirundinis swallow barn France 85 HIR1 HIR1B Co_39 FM179366 Tro_44 + Tro_44 FN257844 HOM. Barn bird nest - from D. hirundinis swallow barn France 85 HIR6 HIR6a Co_38 xxxxxxxx Tro_44 +Tro_44 FN257845 HOM. Barn bird nest - from D. hirundinis swallow barn France 85 HIR6 HIR6b Co_39 xxxxxxxx Tro_44 +Tro_44 FN257849 HOM. Barn bird nest - from D. hirundinis swallow barn France 85 HIR6 HIR6c Co_39 xxxxxxxx Tro_44 + Tro_44 FN257848 HOM. Barn bird nest - from id. D. hirundinis swallow barn 18 LB18c LB18c Tro_44 + Tro_44 FN257846 HOM. Barn bird nest - D. hirundinis swallow building roof France 38 OC OC5 Co_38 AM921862 Tro_44 + Tro_44 FN257846 HOM. Barn bird nest - id. D. hirundinis swallow building roof France 38 OC OC6 Co_38 AM921862 Tro_44 + Tro_44 FN257847 HOM.
D. hirundinis Tit layer farm USA PARATRPARATR8 Co_37 FM208746 Tro_43 + Tro_42 FN257851 FN257852 Barn bird nest - from D. hirundinis swallow barn France 69 TB081 TB082 Co_38 xxxxxxxx Barn bird nest - from D. hirundinis swallow barn France 69 TB081 TB083 Co_38 xxxxxxxx Barn bird nest - from D. hirundinis swallow barn France 69 TB081 TB084 Co_38 xxxxxxxx Tro_44 +Tro_44 FN257879 HOM. Barn bird nest - from D. hirundinis swallow barn France 69 TB081 TB085 Co_38 xxxxxxxx Tro_44 + Tro_44 FN257880 HOM. Barn bird nest - from id. D. hirundinis swallow barn France 69 TB085 TB085e Co_56 xxxxxxxx Tro_44 +Tro_44 FN257846 HOM. Barn bird nest - from id. D. hirundinis swallow barn France 69 TB085 TB085g Co_56 xxxxxxxx Tro_44 +Tro_44 FN257846 HOM.
D. hirundinis Troglodyte bird nest USA TROAEDTROAED2 Co_37 FM208747 Tro_42 + Tro_42 FN257853 HOM.
190 Publication V
bird nest - nest D. longipes Tit box France 69 ENVL083 ENVL083a Co_40 FM179365 Tro_48 + Tro_48 FN257854 HOM. bird nest - nest id. D. longipes Tit box France 69 ENVL083 ENVL083b Co_40 FM179365 Tro_47 + Tro_47 FN257855 HOM. bird nest - nest D. longipes Tit box France 69 ENVL088 ENVL088a Co_41 FM208743 Tro_46 + Tro_46 FN257856 HOM. bird nest - nest id. D. longipes Tit box France 69 ENVL088 ENVL088b Co_41 FM208743 Tro_45 + Tro_45 FN257857 HOM. bird nest - nest D. longipes Sparrow box France 13 JBO108 JBO108a Co_42 xxxxxxxx Tro_41 + Tro_41 FN257837 HOM. bird nest - nest D. longipes Sparrow box France 13 JBO49 JBO495 Co_42 AM921869 Tro_41 + Tro_41 FN257858 HOM. bird nest - nest id. D. longipes Sparrow box France 13 JBO49 JBO49DL2 Co_42 AM921869 Tro_41 + Tro_41 FM897376 HOM. bird nest - nest id. D. longipes Sparrow box France 13 JBO49 JBO49DL3 Co_42 AM921869 Tro_41 + Tro_41 FN257860 HOM. D. quintus USA ADqui ADqui Co_Dquintus AM921882
O. bacoti lab strain France 75 Ob Ob Co_Obacoti FM179677 Tro_Obacoti + Tro_Obacoti FN257767 HOM.
T. pyri lab strain France 34 T_pyri T_pyri Co_Tpyri FM179364 Tro_Tpyri + Tro_Tpyri xxxxxxxx HOM.
191 Publication V
Appendix 2. Information about the matrix of encoded In/Del (Tropomyosin)
List of discrete characters encoded from insertions/deletions recorded along the Tropomyosin sequence On the whole dataset, 12 different regions with insertions/deletions variable intraspecifically have been identified based on all Tropomyosin gene copies obtained from individuals belonging to D. gallinae (named using the first letters of alphabet, followed in some cases by a number). Additionnally, insertions/deletions which were not potentially discriminant between populations of D. gallinae were named using ID (In/Del) followed by a number.
Character Presence of Position of insert Character state Code Character state species of Dermanyssus Comments name intraspecific (based on concerned variation alignment ISOL- according to TRO1) presence/absence of the considered insert ID1 37 c 1 D. carpathicus absent 0 all others ID2 106-108 ttc 2 D. gallinae gtc 1 D. apodis absent 0 D. carpathicus, D. hirundinis, D. longipes A x 130-132 gtg 1 D. gallinae absent 0 D. gallinae, D. apodis, D. carpathicus, D. hirundinis, D. longipes ID3 x 170 g 1 D. longipes (population ENVL08) absent 0 D. gallinae, D. apodis, D. carpathicus, D. hirundinis, D. longipes B1 x 179-184 ttgtct 1 D. gallinae g 0 D. gallinae, D. apodis, D. hirundinis, D. longipes tag 2 D. carpathicus B2 x 195-199 ctttg 1 D. gallinae, D. apodis ttttt 2 D. gallinae cttta 3 D. gallinae tttcg 4 D. carpathicus, D. hirundinis, D. longipes ttttg 5 D. gallinae
192 Publication V
absent 0 D. gallinae ID4 x 220-223 aaag 1 D. hirundinis, D. longipes (+ 1 D. gallinae individual 8018) aaaa 2 D. carpathicus gaag 3 D. gallinae, D. apodis absent 0 D. longipes (population ENVL08) ID5 228-229 tt 1 D. carpathicus absent 0 D. gallinae, D. apodis, D. hirundinis, D. longipes ID6-C1 x 230-254 absent 0 D. longipes (population ENVL08) complex c/tggtttgaaccgaa 2 D. carpathicus, D. hirundinis, D. longipes /gtttgaatt tggtttgaaccgaaaa 3 D. apodis numerous agtttag nucleotide mutations tggc/tg/ttgaaccg 3 D. gallinae gaaaat/ag/ttgaa tggcgtgaa 1 D. gallinae ID7-C2 x 258-267 a/ctttttaaaa 2 D. gallinae complex atttttttta 2 D. gallinae absent 0 D. gallinae atgtttaaaa 2 D. apodis gttttttaaa 2 D. longipes (population ENVL08) gtttttaaa 1 D. longipes (population ENVL08) gttttttaaaa 2 D. hirundinis, D. longipes gttttaaatt 2 D. carpathicus ID8 x 273 c 1 D. carpathicus absent 0 D. gallinae, D. apodis, D. carpathicus, D. hirundinis, D. longipes ID9 286-288 cta 2 D. gallinae, D. apodis cca 1 D. gallinae absent 0 D. carpathicus, D. hirundinis, D. longipes ID10 297-299 ttc 1 D. hirundinis absent 0 D. gallinae, D. apodis, D. carpathicus, D. longipes
193 Publication V
D x 318-322 tagta 1 D. gallinae, D. apodis, D. carpathicus, D. hirundinis, D. longipes absent 0 D. gallinae E x 347-353 cgctcga 1 D. gallinae, D. apodis, D. hirundinis, D. longipes tgctcga 2 D. carpathicus absent 0 D. gallinae ID11 354-355 aa 1 D. gallinae, D. carpathicus, D. hirundinis, D. longipes ga 2 D. apodis absent 0 D. longipes (population ENVL08) ID12 377-380 atac 2 D. apodis a 1 D. gallinae, D. carpathicus absent 0 D. hirundinis, D. longipes F x 398-408 attggacc 1 D. gallinae, D. apodis attggact 5 D. longipes atcggat 2 D. carpathicus attggaccgac 4 D. hirundinis attggaccgc 3 D. longipes absent 0 D. gallinae ID13 x 415-416 cc 1 D. gallinae, D. apodis, D. carpathicus, D. hirundinis, D. longipes c 2 D. hirundinis absent 0 D. carpathicus G x 422-425 gtca 1 D. gallinae, D. apodis, D. carpathicus, D. hirundinis, D. longipes gcca 2 D. gallinae gtcc 3 D. gallinae gcct 4 D. gallinae absent 0 D. gallinae H x 426-433 ggcggc 1 D. gallinae ggcggctc 2 D. apodis absent 0 D. gallinae, D. carpathicus, D. hirundinis, D. longipes ID14 x 445-449 tgaag 1 D. carpathicus tgaaa 2 D. carpathicus c 3 D. gallinae (one individual) absent 0 D. gallinae, D. apodis, D. hirundinis, D. longipes
194 Publication V
ID15 453-455 ctg 1 D. gallinae, D. apodis absent 0 D. carpathicus, D. hirundinis, D. longipes ID16 464-467 agct 1 D. apodis absent 0 D. gallinae, D. carpathicus, D. hirundinis, D. longipes ID17 483 g 1 D. gallinae, D. apodis, D. carpathicus, D. hirundinis, D. longipes absent 0 D. hirundinis ID18 506-508 atg 1 D. gallinae, D. apodis absent 0 D. carpathicus, D. hirundinis, D. longipes J x 564-582 TGAx2 1 D. gallinae TGC instead of TGA in some cases TGAx3 2 D. gallinae TGC instead of TGA in some cases TGAx4 3 D. gallinae TGC instead of TGA in some cases TGAx5 4 D. gallinae TGC instead of TGA in some cases TGAx6 5 D. gallinae TGC instead of TGA in some cases TGAx3 + 6 D. apodis CGGA absent 0 D. carpathicus, D. hirundinis, D. longipes ID19 599-601 tcg 1 D. gallinae, D. apodis absent 0 D. carpathicus, D. hirundinis, D. longipes ID20 636 g 1 D. gallinae absent 0 D. apodis, D. carpathicus, D. hirundinis, D. longipes ID21 639-340 tg 1 D. gallinae, D. apodis, D. longipes cg 2 D. hirundinis absent 0 D. carpathicus ID22 x 684 a 1 D. gallinae
195 Publication V
absent 0 D. gallinae, D. apodis, D. carpathicus, D. hirundinis, D. longipes ID23 688-690 gca 1 D. gallinae, D. apodis absent 0 D. carpathicus, D. hirundinis, D. longipes
196 Publication V
Matrix of encoded In/Del
Character no 111111111122222222223333 Haplotype no 123456789012345678901234567890123 Tro_1 010011103201011111110101011411101 Tro_2 010001103201011111011101011311101 Tro_3 010011103201011111011101011311101 Tro_4 010011103201011111111101011411101 Tro_5 011011103201011111111101011411101 Tro_6 011011103201011111111101011411101 Tro_7 010000103201011111011101011311101 Tro_8 010011103201011111111101011311101 Tro_9 010011103201011111111101011311101 Tro_10 010011103201011111111101011311101 Tro_11 010011103201011111111101011311101 Tro_12 010011103201011111111101011311101 Tro_13 010011103201011111111101011311101 Tro_14 010011103201011111111101011311101 Tro_15 010011103201011111111101011311101 Tro_16 010011103201011111111101011211101 Tro_17 010011103201011111111101011311101 Tro_18 010001103201011111111101011311101 Tro_19 010001103201011111111101011311101 Tro_20 010001103201011111111101011311101 Tro_21 010011103201011111110101011311101 Tro_22 011011103201010111111101011311101 Tro_23 010001103201001111112001011211101 Tro_24 011011103201011111111101011411101 Tro_25 010001103201001111112001011211101 Tro_26 010011103201011111010101011311101 Tro_27 011011101201011111111101011511101 Tro_28 020001103211011212111301111610101 Tro_29 020001103211011212111201111610101 Tro_30 100021112200012101111010010000000 Tro_31 100021112200012101111010010000000
197 Publication V
Tro_32 100021112200012101111010010000000 Tro_33 100021112200012101111010010000000 Tro_34 100021112200012101111010010000000 Tro_35 100021112200012101111010010000000 Tro_36 100021112200012101111010010000000 Tro_37 100021112200012101111010010000000 Tro_38 100021112200012101111010010000000 Tro_39 100021112200012101111010010000000 Tro_40 100021112200012101101010010000000 Tro_41 000001102220011110113000010000100 Tro_42 000001102220111110121000010000200 Tro_43 000001102220111110121000010000200 Tro_44 000001102220111110111000000000200 Tro_45 000101000120011010111000010000100 Tro_46 000101000120011010111000010000100 Tro_47 000101000120011010111000010000100 Tro_48 000101000220011010111000010000100 Tro_49 011011103201011111111101011411101 Tro_50 011011103201011111111101011411101 Tro_51 011011103201011111111101011411101 Tro_52 011011103201001111112001011111101 Tro_53 010001103201001111114001011111101 Tro_54 011011103201011111111101011511101 Tro_55 010000103201011111011101011311101 Tro_56 010001103001011111011101011311101 Tro_57 010000103201001111112001011211101 Tro_58 011011103201001111112001011111101 Tro_59 010011103201001111112001011211111 Tro_60 011011103201001111112031011111101 Tro_D._hirsutus 100001102221011110111300011100101 Tro_T._pyri 000131?0002001011000000000100???? Tro_O._bacoti 100031?0040002111??1?01001000????
198 Publication V
Appendix 3. ABC analysis of case II. 30 clusters of the 180 branching tree histories when considering 5 populations cladograms.
A B CDE
0 1 2 34 0 2 1 3 4 0 3 1 2 4 0 4 1 2 3 1 2 0 34
F G H I J
1 3 0 2 4 1 4 0 2 3 2 3 0 14 2 4 0 13 3 4 0 1 2
K L M N O
0 2 0 34 1 3 0 2 4 1 4 0 2 3 0 2 1 34 0 3 1 2 4
P{ Q R S T
0 4 1 2 3 0 1 2 3 4 0 3 2 14 0 4 2 13 0 1 3 2 4
U V W X Y
0 2 3 14 0 4 3 1 2 0 1 4 2 3 0 2 4 13 0 3 4 1 2
Z AA AB AC AD
0 1 2 34 1 0 2 3 4 2 0 1 34 3 0 1 2 4 4 0 1 2 3 199 Publication V
Appendix 4. Prior distribution and Bayesian probabilities of different scenarios in the three studies using an ABC method. A. Case I (intraspecific, D. gallinae, host type). B. Case II (intraspecific, D. gallinae, individual isolates, with diverse geographica locations). C. Case III (interspecific, D. gallinae, D. hirundinis, D. apodis).
A. Case I (intraspecific, D. gallinae, host type)
Prior distributions used for the intra-specific population genetics analyses of D. gallinae. Symbol description Prior distribution Demographic parameters Ne Population sizes Uniform(0, 500000) t Splitting times Uniform(0, 2.5Mya) m Migration rates Uniform(0, 0.01) Mutations parameters COI mutation rate for COI locus Lognormal(-5.00, 0.06) Tropo mutation rate for Troposymin locus Lognormal(-5.68, 0.15)] both mutation rate for both locus Normal(6.92E-6, 2.95E-6)
Bayesian probabilities of different scenarios regarding presence or absence of migration between D. gallinae groups with different host types. topologies both loci 1 migration 3.6% 2 migration 3.6% 3 migration 3.7% 4 migration 4.0% 5 migration 4.4% 6 migration 3.5% 7 migration 4.0% 8 migration 3.4% 9 migration 3.8% 10 migration 4.0% 11 migration 3.6%
200 Publication V
12 migration 3.9% 13 migration 3.8% 14 migration 3.8% 15 migration 3.3% 16 migration 3.7% 17 migration 3.6% 18 migration 3.6% no 2.2% 19 migration no 2.1% 20 migration no 2.1% 21 migration no 2.7% 22 migration no 1.9% 23 migration no 2.0% 24 migration no 1.8% 25 migration no 1.9% 26 migration no 1.8% 27 migration no 1.8% 28 migration no 1.1% 29 migration no 1.5% 30 migration 31 no 1.9%
201 Publication V
migration no 1.8% 32 migration no 1.5% 33 migration no 1.4% 34 migration no 1.5% 35 migration no 1.9% 36 migration
Bayesian probabilities of different scenarios regarding population branching tree histories between D. gallinae groups with different host types. topologies both loci 1 0% 2 0% 3 0% 4 0% 590% 6 0% 7 0% 8 0% 9 0% 10 0% 11 0% 12 0% 13 0% 14 0% 15 0% 16 0% 17 0% 18 9%
202 Publication V
B. Case II (intraspecific, D. gallinae, individual isolates, with diverse geographica locations).
Bayesian probabilities of different clusters of population branching tree histories between D. gallinae populations with different geographical location. Clusters both loci A 0% B 4% C 1% D 16% E 0% F 0% G 0% H 0% I 0% J 0% K 0% L 0% M 0% N 0% O 0% P 0% Q 44% R 0% S 0% T 0% U 34% V 0% W 0% X 0% Y 0%
203 Publication V
Z 0% AA 0% AB 0% AC 0% AD 0%
Bayesian probabilities of different population branching tree histories from the most supported clusters between D. gallinae populations with different geographical location. topologies both 19 0.00 22 0.00 25 0.02 27 0.00 28 0.00 46 0.00 55 0.00 56 0.01 60 0.00 63 0.00 64 0.24 73 0.00 130 0.00 133 0.00 137 0.00 141 0.72 157 0.00 164 0.00
C. Case III (interspecific, D. gallinae, D. hirundinis, D. apodis)
Prior distributions used for the inter-specific population genetics analyses of Dermanyssus. Symbol description Prior distribution
204 Publication V
Demographic parameters Ne Population sizes Uniform(0, 500000) t Splitting times Uniform(0, 10Mya) m Migration rates No migration considered Mutations parameters COI mutation rate for COI locus Lognormal(-5.00, 0.06) Tropo mutation rate for Troposymin locus Lognormal(-5.68, 0.15)] both mutation rate for both locus Normal(6.92E-6, 2.95E-6)
Bayesian probabilities of different scenarios regarding population branching tree histories between different Dermanyssus species. topologies both loci 1 35% 2 10% 354%
205 MUTATIONS InDel (Dermanyssus dataset) InDel (D. gallinae dataset)
1 51 101 151 201 251 301 351 401 451 501 551 601 651 701 751 Nucleotide position on the basis of alignment ISOL_TRO1
Figure 1
D. gallinae
D. apodis
L1
D. hirundinis
A.
D. gallinae
L1
D. apodis
French layer farms Non French layer farms Chicken farms Amateur layer house B. Pet birds
Wild Figure 2 avifauna
1 60
50
40 mutated positions 30 deleted sequences added sequences 20
10
0 Number of observed pure synapomorphies pure observed of Number
s ) ) inis nae AS hicu li pes P VL08 lade A nae L1 rund gal /c ngipes) N li D apodis B o pes E arpat hi D longipes D longipes ngi c D ;D pes D lade s;D longi s lo D gal C u D ngi lo D (D apodi arpathic apodis;D carpathicus)(D apodis;D hirundinis) (D hirundinis;D c l (D (D Observed entities
Figure 3
D. longipes EN D. longipes PAS D. hirundinis
D. carpathicus
D. apodis α D. gallinae s. str. part
D. gallinae L1
D. gallinae s. str. part
Figure 4 A.
2 D. longipes PAS D. hirundinis
D. longipes EN
D. carpathicus
D. apodis α D. gallinae s. str. part
D. gallinae L1
D. gallinae s. str. part
Figure 4 B.
D. longipes EN
D. longipes PAS D. hirundinis
D. carpathicus
D. apodis
α
D. gallinae L1
D. gallinae s. str.
Figure 4 C.
3 Tro Tpyri Tro Obacoti Tro 41 D. longipes PAS Tro 44 0.59 Tro 42 D. hirundinis 0.73 Tro 43 Tr o 48 0.93 Tro 47 Tro 45 D. longipes EN 0.83 Tro 46 1.00 0.81 Tro 30 Tro 32 1.00 Tro 31 Tro 33 0.51 Tro 40 D. carpathicus Tro 37 0.79 Tro 38 0.96 Tro 39 Tro 36 0.73 0.99 Tro 34 1.00 Tro 35 Tro Dhirsutus 1.00 Tro 28 Tro 29 D. apodis 0.88 Tro 26 1.00 α Tro 3 Tro 21 Tro 15 0.91 Tro 11 Tr o 14 Tro 13 D. gallinae L1 Tr o 12 Tr o 8 0.95 Tr o 10 Tr o 9 Tro 16 Tro 17 Tro 2 Tro 55 Tro 20 0.72 D. gallinae 1.00 Tro 19 Tro 18 Tr o 7 0.1 0.96 Tro 56 Tro 52 1.00 Tro 58 . 1.00 Tro 60 Tro 59 0.87 Tro 53 0.74 Tro 57 0.73 Tro 23 0.65 1.00 Tro 25 Tro 22 Tro 6 Tr o 24 Tr o 1 . 0.96 Tr o 4 Tro 5 . Tro 51 Figure 4 0.96 Tro 27 0.89 Tro 54 0.97 Tr o 49 D. 1.00 Tr o 50
D. hirundinis D. longipes EN D. longipes PAS
D. carpathicus
D. apodis
D. gallinae L1
Lmt2
D. gallinae s. str.
Lmt3
Lmt1 E.
4 D. apodis
D. longipes PAS D. Longipes EN
D. hirundinis
D. carpathicus
D. gallinae s. str. part
D. gallinae L1
Lmt1
Lmt2
D. gallinae s. str. part
Lmt3 Figure 4 F.
Top5: Top141:
French Non-hen Wild birds Non-French 8006 IL 8020SK JBO farm farm farm
A. Top3: B.
Figure 5 D. gallinae D. hirundis D. apodis C.
5 5.3 Arthropodofaune de nids d’oiseaux en agroécosystème et implication des Dermanyssoidea hématophages : publication VI
a - Présentation Après des études mettant à profit des outils phylogénétiques et de génétique des populations, certains éléments de l’écologie des espèces françaises du groupe gallinae ont été partiellement clarifiés. Ou plutôt, certaines questions plus précises ont pu être formulées, sur la base d’indices évolutifs. En particulier, les éléments qui président à l’adaptation accrue de D. gallinae aux habitats modifiés par l’homme tels les élevages, comparé aux autres espèces de Dermanyssus, demeurent pour l’heure très hypothétiques (cf. éléments des résultats 3 de la publication IV, énumérés dans le § 5.1.a.3). Parmi ces hypothèses est évoqué le rôle de la composition des communautés d’arthropodes peuplant les nids d’oiseaux. Les agroécosystèmes sont en général caractérisés par une altération de la faune, si l’on compare avec la plupart des écosystèmes naturels. La biodiversité dans divers groupes zoologiques est souvent modifiée, en particulier chez les arthropodes et les oiseaux. En outre, la prise en compte de la préservation de l’environnement et le souci d’une agriculture durable nécessitent de limiter au maximum l’utilisation de produits phytosanitaires. Ainsi plusieurs types d’agroécosystèmes sont-ils générés par la diversification des méthodes de lutte. J.C. Bouvier (INRA Avignon) compare ainsi depuis plusieurs années nichées et couvées d’oiseaux obtenues dans des nichoirs placés dans des parcelles soumises à trois types différents de modes de lutte contre les arthropodes déprédateurs des fruitiers considérés : conventionnel, biologique, intégré. Des nids de mésange prélevés in natura viennent compléter l’étude. Les nids d’oiseaux constituant par excellence des îlots, leur arthropodofaune est tout à fait particulière et riche. En effet, de par leur isolement naturel et la présence du vertébré, ces îlots offrent une biodiversité souvent remarquable in natura : un microécosystème plus ou moins équilibré se crée, impliquant un guilde de détritivores et saprophages souvent riche, se nourrissant sur les déchets produits/introduits par le vertébré, ainsi qu’une guilde de prédateurs tout à fait intéressante, se nourrissant des ectoparasites apportés par le vertébré et des détritivores et saprophages sus-cités. L’isolement caractéristique de ce type d’habitat entraîne un certain degré de spécialisation chez ces arthropodes, certains parmi l’une ou l’autre des trois guildes principales (détritivores/saprophages, prédateurs, parasites) étant strictement inféodés aux nids d’oiseaux. Afin d’obtenir un aperçu du microécosystème de cet acarien chez des oiseaux particulièrement communs et de détecter une éventuelle corrélation entre arthropode prédateur et présence de D. gallinae, entre pesticides et présence de D. gallinae, nous avons engagé, en collaboration avec l’INRA d’Avignon, un étude comparative de l’arthropodofaune de nids d’oiseaux en agroécosystèmes de vergers.
5.3.a.1 Objectifs Les objectifs étaient scindés en deux axes : L’arthropodofaune particulière des îlots constitués par les nids d’oiseaux dans les agroécosystèmes est-elle aussi affectée par les pratiques agricoles, malgré leur isolement ? Comment sont affectés les mésostigmates hématophages inféodés aux oiseaux dans les agroécosystèmes de verger ?
211 5.3.a.2 Matériel et méthodes L’étude est venue s’intégrer à une étude déjà très avancée, menée par J.C. Bouvier, sur la biodiversité de l’avifaune dans ce type de milieu et les impacts des pratiques agricoles sur divers paramètres afférents aux oiseaux (biodiversité, fitness, …). Elle reposait sur la comparaison de cette faune particulière observée dans trois grands types de vergers de fruitiers (fruits à pépins) (ainsi qu’in natura) : lutte biologique, lutte chimique et lutte intégrée. L’arthropodofaune a été appréhendée sous trois volets plus ou moins approfondis : (1) la simple présence ou non d’arthropodes dans les nids analysés d’une manière standardisée, (2) la diversité des arthropodes isolés, (3) la nature de certains ectoparasites hématophages (Acari : Mesostigmata) parmi ces arthropodes. Les résultats présentés ici sont issus des deux années d’échantillonnage (2007 et 2008). Des tests statistiques ont évalué le degré de significativité des différences observées dans les trois volets.
5.3.a.3 Principaux résultats Le ratio du nombre de nids ayant permis de détecter des arthropodes par rapport au nombre de nids « vides » (volet n°1) s’est avèré significativement marqué par les pratiques de lutte : les nids « vides » étaient nettement plus nombreux en verger conventionnel. En revanche, les résultats des volets n°2 et 3 ont mis en évidence des tendances, mais n’ont pas montré, pour la plupart, de significativité notoire, sans doute du fait de la trop faible occurrence des différents groupes d’arthropodes pour l’obtention de données statistiquement fiables. La comparaison de la biodiversité relevée entre nids de vergers et nids prélevés in natura a suggèré une variation en accord avec les résultats d’Ives et al. (2000). Ces auteurs ont démontré que la richesse en espèce augmente la stabilité au niveau de la communauté dans la mesure où ladite communauté fournit des espèces qui ont des chances d’être tolérantes à différentes fluctuations environnementales, et ainsi sont complémentaires des autres. Ici, les acariformes détritivores étaient faiblement diversifiés et leur nombre a semblé influencé par les pratiques agricoles, alors que les ectoparasites considérés comportent des espèces qui ont été démontrées comme très différentes en terme de tolérance aux environnements modifiés par l’homme et leur nombre se trouve ne pas être affecté. En revanche, les psocoptères détritivores se sont avèrés représentés par une seule espèce, Liposcelis bostrychophila Badonnel, 1931, apparemment non influencée par les modalités de lutte. Mais il s’agit précisément d’une espèce qui a fait la preuve d’une capacité d’adaptation remarquable dans des environnements modelés par l’homme, développant rapidement des résistances croisées multiples. Quant au genre Dermanyssus, D. gallinae s’est montré une fois de plus différent des autres en terme de tolérance à des pratiques humaines, puisqu’il a été isolé dans les vergers et non in natura chez les mésanges du genre Parus. En l’occurrence, le seul point commun apparent entre verger et élevage semble être les actions de contrôle des arthropodes (et de manière corrélée, la présence réduite de prédateurs). Toutefois, D. gallinae n’était présent qu’en verger biologique ou intégré, jamais en verger conventionnel. Il semblerait peu tolérant aux traitements rencontrés dans les vergers conventionnels. Un acarien d’une autre famille, O. sylviarum (Macronyssidae) a semblé, par ailleurs, présenter un spectre de tolérance plus large encore puisqu’il était aussi présent en verger conventionnel, et de manière récurrente. Notons enfin qu’O. sylviarum, fréquemment rencontré dans les nids d’oiseaux testés, demeure aujourd’hui absent des élevages de pondeuses en France. Cette espèce est un ravageur notoire en pondeuses aux Etats-Unis. A l’inverse, D. gallinae, le pou rouge des volailles, présent in natura dans les deux continents, est un ravageur notoire en France, pas aux Etats-Unis. Par ailleurs, D. gallinae bien que présent dans 80% des élevages de pondeuses n’a jamais été recensé dans les nids de mésange charbonnière prélevés in natura. Une analyse comparative de la génétique des populations
212 de chacune des deux espèces en France et aux USA (en élevages comme in natura) pourrait apporter d’intéressantes informations quant à l’impact différentiel des pratiques d’élevages des deux pays.
213
Publication VI
Arthropodofauna in bird nests as an indicator for agricultural practices' impact in pear and apple orchards
Roy L.*, Bouvier J.C.**, Lavigne C.**, Galès M.**, Chauve C.M.* and Buronfosse T.* [email protected] * Université de Lyon - Ecole Nationale Vétérinaire de Lyon – Laboratoire de Parasitologie – 1 av. Bourgelat – 69280 MARCY L’ETOILE - FRANCE **INRA - Unité PSH - Site Agroparc – 84914 AVIGNON Cédex 9 - FRANCE
Abstract Man-induced environmental alterations may be particularly important in agroecosystems. Recent changes in agriculture in the interest of the sustainable development may have an impact on these alterations. The aim of present study is to estimate whether the particular arthropodofauna of bird nests in agroecosystems is affected by three types of agricultural practices (organic, conventional and integrated control) in an orchard agroecosystem and how bird parasites respond to the management practices. The amount and diversity of arthropodan communities developping in bird nests in the context of pear and apple orchards was evaluated using standardized nest boxes. A few additional samples from natural habitats have also been analyzed and obtained results compared. A comparison of observed nest arthropodofauna according to the pest control methods allowed detecting some strong impact of synthetic pesticides on non target arthropoda which are enclosed within the island ecosystem of nest. Thus, statistically more arthropod-free nests were recorded in conventional (chemically controlled) orchards than in integrated and still more than in organic orchards. Due to the weak number of occurrences, results on arthropod biodiversity were not as significant as the simple notation of presence/absence. Nevertheless, following tendencies were evidenced: 1) coleoptera were more numerous and more diverse in natura vs in orchards, even considering only organic orchards. 2) some opportunistic arthropods which were encountered are also pests for human in stored food or farms, either in the guild of saprophagous/detritivorous arthropods, or in the guild of ectoparasites. And yet, some of them did not show any clear sensitivity to chemical control methods, as they were recurrently found in conventional orchards. This was not unexpected as these arthropods have already shown an enhanced flexibility. Finally, among these less sensitive and commonly pest arthropods, Ornithonyssus sylviarum, the Northern Fowl Mite was recorded. It is remarkable that O. sylviarum is so commonly encountered in bird nests under test, yet absent from French layer farms, whereas it is a serious pest in layer farms in the USA. The opposite applies to Dermanyssus gallinae, the Poultry Red Mite, an important layer pest in Europe. This suggests that different farming practices between both continents might explain the paradox of the omnipresence of both species in wild avifauna and a selective presence in layer farms.
Key words arthropodofauna, bird nest, ectoparasite, Ornithonyssus, Dermanyssus
214 Publication VI
1. Introduction 215 2. Material and methods 216 2.1. Sampled orchard nests 216 2.1.1. The study areas 216 2.1.2. Nest boxes and birds 217 2.1.3. Natural environments 217 2.1.4. Nests analysis 217 2.1.5. Presence/absence notation 217 2.1.6. Arthropoda identifications 217 3. Results 218 3.1. Presence/absence of arthropoda 218 3.2. Diversity of arthropoda 219 4. Discussion 221 4.1. Presence/absence of arthropoda 221 4.2. Diversity of arthropoda 221 4.2.1. Ecological role of isolated taxa within nests’ arthropoda communities 221 4.2.2. Natura vs orchard nests 223 4.2.3. Hematophagous Mesostigmata 225 5. Conclusion 225 6. Acknowledgments 226 7. References 226 8. Legends 229 9. Appendix 1. Host and location information for tit nests sampled in natura. 230 10. Appendix 2. Mite species, sample location and EMBL accession numbers for sequenced samples 231
1. Introduction
Agroecosystems are usually characterized by a fauna that is different from most of natural ecosystems. Biodiversity in various zoological groups is often altered, especially in arthropods and birds. Nevertheless, the need of preservation of environment and practicing sustainable agriculture requires maximum restriction of phytosanitary products’ use. Thus, several types of agroecosystems have been generated by the diversification of control methods. Numerous studies have been carried out on the impact of development and/or phytosanitary treatments in agroecosystems on abundance or diversity of the entomofauna sensu lato [1, 2, 3, 4, 5]. Nevertheless, most of these studies focus on crops auxiliary arthropods. On the other hand, the bird diversity in orchards changes with the pest control method according to several studies [6, 7, 8]. And yet, bird nests constitute islands par excellence, in which a particular and rich arthropodofauna is to be found. Indeed, due to their natural isolation and to the presence of a vertebrate, these islands are provided with an often amazing diversity of insects and arachnids in natura [9, 10, 11, 12, 13, 14, 15]: a rich microecosystem
215 Publication VI is created in many bird nests. It involves a large guild of detritivorous or saprophagous arthropods, which feed on droppings and/or wastes produced/introduced by the bird inhabitant. As well a guild of predator insects and arachnids is often present, which feeds on ectoparasites (carried by the vertebrate host) and/or on detritivorous or saprophagous arthropods cited above. The particular isolation of such habitats may result in more or less nest-specialized arthropods in each guild. As a result, some of them, in any of these three guilds are strictly confined to bird nests. Parasites are usually at least specialized on birds in general, if not on the bird species under consideration. But even in detritivorous/saprophagous and predator guilds, some arthropods are specific to bird nests. Moreover, the regulating action of some of these arthropods may play a role in the control of parasite development. Thus, the proliferation of some ectoparasites in farms might be a side effect of farming practices. The aim of present study was to address following questions: is the particular arthropodofauna of bird nests from nest boxes in agroecosystems also affected by agricultural practices ? How are bird parasites affected by the management practices in an orchard agroecosystem ? In the present study, the arthropodofauna of nest boxes sampled in three different types of apples and pear orchards were sampled and compared. These three types were defined by the control methods used against crop pests: orchards with organic control, chemical control or integrated control. An additional dataset was obtained from nests of birds sampled in natura. 2. Material and methods
2.1. Sampled orchard nests
2.1.1. The study areas The study area is located in southeastern France in the Avignon region (43°96’ N, 4°82’ E). Studied orchards were commercial apple and pear orchards located on privately owned farms. Among the orchards followed, 5 were conducted under organic, 5 under conventional and 5 under Integrated Pest Management (IPM) strategies for each fruit species. In organic orchards, the use of synthetic chemicals (both fertilisers and pesticides) is excluded according to the commission regulation (European Community 473/2002) that amended the council regulation (European Economic Community 2092/91). Conventional orchards were managed using synthetic chemical pesticides only according to the 1997 French national charter of apple production. The IPM orchards were conventional orchards using mating disruption against Cydia pomonella, the main insect pest. These orchards displayed a similar pattern of a cultivated area restricted to one hectare inside a larger orchard unit and is surrounded by single-rowed hedgerows used for protection against the north prevailing wind. All studied apple orchards were surrounded only by orchards conducted under a similar protection strategy. Orchards had also been chosen for their similar pattern in terms of local and landscape features that might influence bird communities. We tested this similarity prior to analyses, and checked in particular that organic, integrated and conventional orchards did not differ in that respect to avoid confounding effects (p > 0.1801). In 2008, in each fruit tree species, the analysis has been conducted in 5 organically controlled orchards, 5 conventionally controlled orchards and 5 controlled using IPM methods. A reduced sampling campaign was also carried out in 2007, with only apple orchards tested.
216 Publication VI
2.1.2. Nest boxes and birds Arthropodan communities in bird nests were collected in nest boxes. The nest boxes were the Schwegler 1B type. Each orchard had five nest boxes installed on 1 ha at least two years before the present study. Each nest box was located on an apple tree 2.5 m from the ground, 30 m from its nearest neighbor, and 20 m from surrounding hedgerows. Entrances of all nests boxes faced southeast to avoid both the north prevailing wind and the south prevailing rain. Birds occupying nest boxes were in most cases great tits (Parus major). In a few nest boxes, blue tits (Parus caeruleus) (1%) or Eurasian Tree Sparrows (Passer montanus) (9%) nestlings were found. Every nest in occupied nest boxes was systematically collected after the young had fledged. Complete nests were removed from the nest box and then stored in a closed plastic bag.
2.1.3. Natural environments Additionally, in order to get an insight into the arthropodofauna to be expected under natural conditions, 21 nests of tits (Parus sp) from nest boxes installed in natura in diverse regions in France (mainly Center or South-eastern France) were sampled in 2007 and analysed using same methods as explained above. Location information is available in Appendix 1.
2.1.4. Nests analysis Arthropoda were isolated from nests following the method described in de Lillo (2001) slightly modified (see Roy et al. 2009b). This method is based on the immersion of the nest in a fixed quantity of water and subsequent filtration using stacked sieves. Next, the obtained filtrate is reimmersed and observed using a stereoscopic magnifying glass. Prior to immersion, a macroscopic observation of the dry sample was systematically done in order to detect larger arthropoda, which might be lost during the filtrate analysis. As compared to Berlese funnel method used in Burtt et al (2001), de Lillo’s method allows detecting both living and dead arthropoda. The arthropodofauna is explored following two different points of view: (1) the simple presence or absence of any arthropod in nests under test is noted, (2) isolated arthropods are identified. This study uses some nest boxes which have been installed in the framework of a study dealing with the impact of agricultural practices on avifauna in agroecosystems.
2.1.5. Presence/absence notation The detection of arthropods led to the following classification of nests: Arthropod-free nest: a nest from which no arthropod has been isolated. Arthropod-poor nest: a nest from which only 1 or 2 individuals belonging to a single primary group (as defined below, in the Results section) have been isolated. Arthropod-rich nest: a nest from which more than 1 primary group or more than 2 individuals of a single primary group have been isolated. In order to get an overview of the number of individuals per primary group and per nest, a rough estimation was noted.
2.1.6. Arthropoda identifications The identification of arthropoda has been performed in two successive stages: (1) A rough identification at a high taxonomic level such as order or family was first performed by using a stereoscopic magnifying glass (2007 and 2008). (2) A species identification was then performed in recurrent taxa, described below (see § Results) (all taxa in 2008, only Coleoptera and Hematophagous Mesostigmata in 2007). Mites:
217 Publication VI
Mesostigmata were identified by LR, acariforms by F. Faraji. Insects: Psocoptera were identified by Z. Kucerova, Coleoptera by R. Allemand. Note: around 2% of isolated arthropods kept undetermined due to bad condition or to inappropriate stage (larval/nymphal stage in some holometabolous insects or in some mites). As for Psocoptera, a sample of isolated insects only has been identified at the specific level (27 individuals, distributed in 6 organic, 5 integrated, 4 conventional orchard nests).
DNA sequencing of some gene portions was performed in hematophagous Mesostigmata following Roy et al. (2009a, b) in order to check specific identity and get some additional population characteristics. In this framework, some mites belonging to Ornithonyssus sylviarum (Canestrini and Fanzago, 1877) (Mesostigmata: Macronyssidae) sampled in layer farms in the USA were kindly provided by Dr. Bradley Mullens and this enabled comparing pest populations with populations from wild bird nests in this species. Phylogenetic analyses were run using Phylo_win 2.0 [16] with the maximum-likelihood (ML) method. All trees were built with 500 bootstrap replicates.
1.1. Statistical tests We tested whether the proportion of empty nests differed among years (qualitative, two levels) and crop protection treatments (qualitative, three levels: organic, integrated and conventional) using a logistic regression (proc GENMOD, SAS 9.01, SAS Institute, Cary, NC, USA) on the binary variable describing whether nests were free of arthropods or not and a logit link function. As nests within orchards are not statistically independent, we introduced a random orchard level in the model. We also tested if the proportion of empty nests differed between nests situated in Organic orchards and in natura in 2007 using logistic regression on the same binary data.
We then compared the average number of primary groups in each nest containing arthropoda (arthropod-rich or arthropod-poor) among years (qualitative, two levels) and crop protection treatments (qualitative, three levels: organic, integrated and conventional) using a generalized linear model assuming a Poisson distribution of the numbers of taxa and a log link function (proc GENMOD, SAS 9.01). As above, we introduced a random orchard level in the model to account for dependence among nests within orchards. We similarly tested if the average number of primary groups differed between nest in Organic orchards and in natura in 2007.
Finally, we tested whether the occurrence of detritivorous Acariforms, Psocoptera and Siphonaptera (fleas) differed between nests situated in Organic orchards and in natura in 2007 also using logistic regression (proc GENMOD, SAS 9.01) on the binary data describing presence or absence of each group in each nest.
In cases where a year effect was significant, analyses were performed using the same independent variables for each study year.
3. Results
3.1. Presence/absence of arthropoda The ratio between arthropod-free /arthropod-rich nests (calculated as a percentage of sampled nests in each year) is similar in both years (P=0.5982) (Fig. 1). In contrast, control methods have a significant impact on this ratio (P=0.0085). An increasing gradient of the number of arthropod-free nests from organic orchards to conventional orchards is conspicuous and
218 Publication VI statistically significant (P=0.0028, organic vs conventional). Nests sampled in natura provided about as many arthropod-free and arthropod-poor nests as nests from Organic orchards (P=0.5450).
3.2. Diversity of arthropoda The occurrence of arthropoda is presented in Fig. 2 distributed in primary groups as defined below. Primary groups represent a consensus between ecological and taxonomical knowledge related to first-glance available morphological characteristics (first stage of identification). Mites were divided into four main groups: hematophagous Mesostigmata, non hematophagous Mesostigmata (predators belonging to Dermanyssoidea, Ascoidea or Eviphidoidea), detritivorous Acariforms (Astigmata), predator Acariforms (Prostigmata: Cheyletoidea). An additional arachnidan group was Aranea (spiders). Most insects were first identified at the order level, with only one division within Hymenoptera: Formicoidea (ants) and parasitoid microhymenopteran was distinguished from each another. Finally, the terrestrial crustacean order Isopoda was sometimes noted. The occurrence of most of arthropodan groups being occasional (if not accidental), the species identification was performed only in repeatedly recovered groups. By far, the best represented groups in orchard nests (t20% occurrence in both years 2007 and 2008) were hematophagous Mesostigmata (Acari), detritivorous/saprophagous Acariforms (Acari) and Psocoptera (Insecta, detritivorous/saprophagous) (Fig. 2). In natura nests, the same groups were repeatedly found, but in addition the following groups: Aranea, Diptera, Coleoptera. As Coleoptera from birds nests may be rather diverse, belong to various guilds, and, in many cases, are more or less nest specific [15, 17], an identification at the species level of adult individuals was also performed in this group (in samples of 2007 and of 2008). A list of species found within these four recurrent primary groups is given in Table 1.
The main noticeable difference concerning taxa between the two years in orchard nests is the occurrence of spiders (31% in 2007, <3% in 2008). The presence/absence of such large arthropodans is likely strongly correlated to the role of nest boxes as shelters from extreme, but transient climatic conditions. Other arthropodan primary groups were represented in accordance to what had been noticed in previous study [18, 19]. A second difference between 2007 and 2008 orchard nests is to be noted in the number of different groups (based on our above described classification) in Organic orchard nests (Fig. 3): the average number of different primary groups in each nest containing arthropoda (arthropod-rich or arthropod-poor) is greater in organic orchards than in integrated and conventional orchards in 2007 (P=0.0139), not in 2008 (P=0.0871). In 2008, a roughly similar number of groups in the three orchard types has been noted. Nests from natural environments that have been sampled the same year, during similar periods, provide significantly more arthropod groups than orchards (P=0.0032), but this is not significant if only Organic orchards are compared to natura (P=0.1469). When considering the specific level in the only two orders of arthropoda identified specifically in both natura and orchards (Coleoptera and Mesostigmata), natura nests contain more coleopteran species than Organic orchards and a similar number of hematophagous Mesostigmata, in the dataset as a whole as well as per nest (cf. Table 2).
Seven species were recurrently collected from orchard nests under test, four of which are hematophagous mesostigmatic mites (D. gallinae (De Geer, 1778), D. carpathicus Zeman, 1979, D. longipes Berlese and Trouessart, 1889, O. sylviarum, Ornithonyssus sp (see Table 1), two of which are the detritivorous prostigmatic mites Hirstia chelidonis Hull, 1931 and Tyrophagus longior (Gervais, 1844), the last one being the well know stored-food pest
219 Publication VI psocopteran insect species Liposcelis bostrychophila Badonnel, 1931 (Table 1). Ten different species of Coleoptera, belonging to four different families, were isolated, eight of which from nests sampled in natura (Table 1). Individuals which were at a larval stage were not identified at the species level. Unfortunately, individuals collected from orchards were mainly larvae (in most cases, a single larva isolated per nest), which we did not identify at the species level, but they all looked like dermestid larvae. Additionally, some other arthropoda have been isolated anecdotically: a diversity of insects, some Isopoda (woodlice) and one Ixodida (ticks), especially in natura nests.
The guild of parasites in nests under test in the present study is rather similar in natura and orchard nests, and not so diverse at the order level. It is almost restricted to Mesostigmata (Acari: Parasitiformes) in the present study except for one orchard nest with Mallophaga in samples of 2008, three orchard nests and three natura nests with fleas in samples of 2007 (Insecta: Siphonaptera) and one natura nest with one hard tick individual (Ixodida). Among the hematophagous Mesostigmata, five species have been collected. In orchard nests of both years, three species of Dermanyssus (Dermanyssidae) and two species of Ornithonyssus (Macronyssidae) have been isolated: D. carpathicus, D. gallinae, D. longipes, O. sylviarum, Ornithonyssus sp (see Table 1). Mites belonging to both famillies have been isolated in 16 % orchard nests under test, 24 % natura nests in 2007 and 27 % orchard nests under test in 2008. Of these five species, the following three ones were isolated from both natura and orchard nests: D. carpathicus, D. longipes, O. sylviarum. D. gallinae and Ornithonyssus sp were only found in orchard nests. In orchards, the three species of genus Dermanyssus were isolated in organic and integrated orchards only, whereas species of genus Ornithonyssus were found in the three types of orchards (Fig. 4). Note that most of individuals of the genus Ornithonyssus belong to O. sylviarum (87.5%). The second detected species of this genus was recorded very occasionally (two nests in Organic orchards in 2008). O. sylviarum and each of the 3 species of genus Dermanyssus were encountered regularly in non-arthropod-free nests in the present study. Of these five species, two are of economic importance in fowl farms, and especially in layer farms: O. sylviarum, in the USA, and D. gallinae, in the Old World. The DNA sequencing of some mitochondrial regions showed that isolates of O. sylviarum from French bird nests under test in present study were not only conspecific to but also very close to an American poultry pest isolate (OSBM ; see Appendix 2 and Fig. 5). Based on an rRNA 16S region isolated following Roy et al [20], isolates sampled in French orchards and isolate OSBM were diverging by 2-3% from each other, and by 1-2 % from an isolate sampled from African Gold Breasted Starlings (Cosmopsarus regius) (accession number AY185362) [21]. Based on an mt-Co1 coding region (Fig. 5), OSBM and isolates sampled in French orchards were diverging by 4% from each other and isolates from orchards were diverging by 1% from two additional French isolates sampled in nests of Montagu's Harrier, Circus pygargus (Linnaeus, 1758) located in a wheat field (FS5 and FS6). Moreover, all French individuals group together as a sister group to the American isolate (OSBM) according to Maximum Likelihood phylogenetic reconstruction, with much shorter branch length between each other, than between their common clade and the close relative O. bacoti (Fig. 5).
On the whole, except for bird ectoparasites, the diversity of arthropoda seems to be slightly higher in natura than in orchards, even organic only. Especially, the species diversity of Coleopteran individuals isolated in natura nests appears greater than in orchard nests (cf. above) and the occurrence percentage of detritivorous Acariforms is higher in natura nests than in orchards (P=0.0172). In contrast, the occurence percentage of Psocoptera is similar (P=0.2414). Several more rarely encountered taxa have also shown a more frequent
220 Publication VI occurrence in natura than in orchards: Siphonaptera (fleas) appeared more often in natura nests, but this had no clear statistical support (P=0.0924), due to the too weak overall amount of occurrences, and Isopoda (woodlice) were found in 12% natura nests and absent in orchard nest. 4. Discussion
4.1. Presence/absence of arthropoda The significant increasing in the ratio of arthropod-free /-poor /-rich nests across the three control methods strongly shows that chemical control has an impact on many non-target arthropoda, which are normally found in bird nests. These differences may not be due to landscape characteristics, as all orchards under test have been shown being similar from a landscape point of view by PCA analyses. It was previously shown by Bouvier et al (2005) that the orchard control methods had a strong impact on orchard avifauna diversity. In contrast, previous studies which reported some differences in soil arthropodofauna of some agroecosystems mainly highlighted the strong impact of landscape and concluded that control methods did not clearly influence the number nor the diversity of arthropods they tested [3, 22]. But both studies dealed with larger and much more mobile insect communities (beetles and butterflies). These arthropods are able to move from one place to another often and very quickly and are not restricted to small island microecosystems as are arthropod communities found in bird nests. The important isolation of nests in nest boxes is not even sufficient however to prevent a strong impact of agricultural practices. Bird nests' arthropod communities are restricted to a fixed place, and may therefore show more precisely the impact of the different control methods on local atmospheric quality than do mobile terrestrial insects. 4.2. Diversity of arthropoda The biodiversity of arthropodofauna in orchard nests under test was not as informative as did the ternary approach (arthropod-free /-poor /-rich nests) concerning the difference between the three control methods in orchards. The different guilds were very less diversified: recurrent detritivorous arthropoda were restricted to two mite and one insect species, predators were almost absent, parasites were mainly represented by five hematophagous mesostigmata species (see Table 1). Even if there are some trends, none is strongly supported (not enough arthropod-rich nests, not enough individuals within arthropod-rich nests). Nevertheless, among these weakly supported trends, some differences can be noticed concerning hematophagous mesostigmatic mites according to control methods (cf. below). In contrast, nests sampled in natura seem to contain more different taxa than orchard nests at least at the species level within Coleoptera. This suggests that even organic methods might have an impact on the biodiversity of arthropodan communities in bird nests. Anyway, nests in natura are not abundant enough, have not been sampled in exactly the same region, nor in comparable environment, so that a firm conclusion cannot be drawn about it.
4.2.1. Ecological role of isolated taxa within nests’ arthropoda communities In the present study, bird nests’ arthropodan taxa may be split into 4 different categories according to the assumed ecological role in nests. Higher taxa and corresponding ecological habits are roughly similar to what Burtt et al [10] found in nest boxes occupied by the Tree Swallow (Tachycineta bicolor: Passeriformes: Hirundinidae), the house wren (Troglodytes aedon: Passeriformes: Certhiidae), the eastern bluebird (Sialia sialis: Passeriformes: Muscicapidae) in North America and to what Krištofík et al [23, 24] found in nests of two
221 Publication VI shrikes (Lanius collurio, L. minor: Passeriformes: Laniidae) and the bearded tit (Panurus biarmicus: Passeriformes: Sylviidae) in Europe. Some differences may be noticed. (1) The occurrence of some arthropoda within bird nests is accidental, due to the tree environment: large saprophagous such as Dermaptera (earwigs) and Isopoda (woodlice) were sometimes isolated, likely due to the shelter role of nest boxes; phytophagous Thysanoptera (thrips) and Hemiptera such as aphids were likely introduced into nests by chance; Embidopsocus enderleini is a psocopteran species found under bark of trees (a single specimen in a single nest); the occurrence of Collembola may be correlated to the large presence of moss in tit nests; the floricolous beetle species Potosia oblonga (Scarabaeidae) was represented by a likely freshly emerged adult (larvae living within tree cavities); two adult individuals of Nalassus dryadophilus (Tenebrionidae) also emerged from tree-living larvae (larvae living under bark trees in this species). Similarly, herbivorous flat bark beetles (Coleoptera: Silvanidae) were noted in many nests by Burtt et al [10]. (2) Some others may, by chance, have found a convenient habitat within nest (adapted to the life within nests, but not specialists). Indeed, larvae of microlepidopteran moths (related to stored-food moths) were found in a few nests likely due to the presence of food substances omitted by chicks. The much more recurrent mite H. chelidonis (and some more rarely encountered T. longior), as well as the insect L. bostrychophila might have found a similar opportunity. These species are not proper to bird nests, but are likely better adapted to such an environment. Formicoidea (ants) are often found in bird nests [10, 25] and some interactions between ants and tits have precisely been shown [26]. Parasitoid microhymenoptera, already recorded in other bird nests by Burtt et al [10], can parasite either lepidoptera larvae or other arthropoda present in the nest. Flies (not parasitic Brachycera Diptera) can find food in droppings and chicks' food. Several species of detritivorous Coleoptera were isolated in natura nests. Attagenus unicolor (Dermestidae) is an unspecific detritivorous insect which can feed on various wastes generated by chicks, Anthrenus pimpinellae, Dermestes undulatus , (Dermestidae) and Alphitobius diaperinus (Tenebrionidae) are feeding on animal substances, which are largely available in insectivorous bird nests (dead insects brought by adult birds but omitted by chicks or dead chicks' corpse). Cryptophagus sp (Cryptophagidae) is a mycetophagous beetle, already noted in some bird nests [27]. These species are not nest specialists and several of them represent common pests in human environments, such as stored food (L. bostrychophila, T. longior), fowl farms (A. diaperinus) and natural history collections (A. unicolor, A. pimpinellae). The presence of A. diaperinus in bird nests has been noted in abundance and in various bird groups [17, 15] and so does L. bostrychophila [28]. (3) Two more or less nest specialist predators were isolated (specialized in bird nest ecosystems): Gnathoncus buyssoni (Histeridae) is a nidicolous predator of sapro- necrophagous larvae and Carcinops 14 striatus (= pumilio) (Histeridae) is an often nidicolous predator of sapro-necrophagous larvae. These species or some congeners are commonly found in raptor nests [14, 15]. They have each been isolated once, the former in natura, the latter in an orchard. The presence of the predator of sapro-necrophagous larvae confirms the family identity of the few dermestid-like larvae isolated from orchard nests, and suggests that one or more of above detritivorous dermestid species are present in orchard nests, but in a much lesser amount than in natura. (4) As for the occurrence of some common bird ectoparasites (specialized in birds), hematophagous Mesostigmata are largely predominant, whereas Mallophaga (shewing lice) and Siphonaptera (fleas) only occur sparsely and rarely. This difference may be correlated to the respective habits of these parasites. Hematophagous Mesostigmata are nidicolous micropredators. In contrast, Mallophaga are spending their complete life on host. Nevertheless, Burtt et al [10] found around six Mallophaga individuals per nest. But their
222 Publication VI study was performed in a different continent (North America) and on different bird species. This may suggest that the prevalence of these parasites is different in these bird species. Krištofík et al [24, 26] did not notice any Mallophaga individuals in shrikes and bearded tits’ nests. As Mallophaga are usually host specific, it is possible that no mallophagan species are used to parasite Parus species. On the other hand, eggs and nymphs, and sometimes adults of Siphonaptera are to be found within nest, which might have been expected more often in nests under test. In natura Burtt et al [10] did not notice any Siphonaptera, whereas Krištofík et al [24, 26] isolated many. This may likely be explained by the divergence in the method for nest analysis more than by the bird species. Indeed, Burtt et al [10] used the Berlese funnel method for extracting arthropoda from collected nests, which requires isolated arthropods to be alive. De Lillo’s method used in present study allows detecting not only living arthropods, but also dead ones. Maybe this could explain the absence of fleas in Burtt et al’s study.
Predators are commonly less numerous than detritivorous/saprophagous and parasites in most ecosystems. Nevertheless, the presence of predators seems to be rather reduced in all tit nests under test if compared to nests of other wild birds. Several different predator Mesostigmata were recurrently isolated in starling nests (Ascidae, Laelapidae, Macrochelidae, Parasitidae,…[29]). Pseudoscorpions are occasionally but regularly found in nests of the barn swallow Hirundo rustica Linnaeus, 1758 (LR pers. observation) or in the bearded tit Panurus biarmicus (Linnaeus, 1758) [24], and regularly in shrikes’ nests (Krištofík et al 2002), among others. And Kristofik et al [24] noted the presence of predator beetles in almost 1/3 nests in P. biarmicus. In present study, only two nests were noted with one predator mesostigmatic mite in each in 2007, one in 2008 and zero in natura nests in 2007. Some rare Cheyletoidea, along with some ants and a few Coleoptera may play this role on different types of prey. This is finally more in accordance with the fauna of predators noted in Burtt et al [10], which is not unexpected as birds under test are all cavity-nesting birds, as opposed to barn swallows, shrikes and bearded tits.
4.2.2. Natura vs orchard nests Although only 21 natura nests have been analyzed and although these samples come from slightly different geographical areas, making results not easily comparable, some elements strongly suggest that diversity within arthropod-rich nests in agroecosystems (whether it is inorganic, integrated or conventional orchard) is not as rich as in natura. Not only detritivorous Acariforms were more frequent and numerous in natura than in orchard nests, but also Siphonaptera, Isopoda and Thysanoptera have been more often (if not only) noted in natura than in orchards. The absence of phytophagous Thysanoptera in orchards may be explained by different tree substrates, but not that of saprophagous Isopoda. This strongly suggests an impact of agroecosystem conditions on these latter arthropods' development. This difference in arthropod diversity and amount between natura and orchard nests is also supported by the specific identity of Coleoptera. Some more or less specific coleopteran insects have been isolated from natura and orchard nests, in a greater number in each nest in natura (3 to >20 per nest) than in orchards (1 to 2). Even compared with Organic orchards, nests sampled in natura are more species rich. This is not exactly in accordance with avian data [6]. These authors demonstrated in the orchards under test that the bird diversity were similar in Organic orchards and in natura. This difference between arthropod and bird diversity is effectively very consistent with the control management, as pests against which organic substances (mineral oil, granulosis virus, sexual pheromones) are used are arthropods,
223 Publication VI not birds. Especially, granulosis virus and sulfur might have a strong impact on non target arthropods living within nests. In contrast, parasitic nidicolous mite species’occurrence is similar in the whole dataset (Table 2). This suggests two possibilities: (1) either mesostigmatic hematophagous mites better tolerate agroecosystem conditions than other encountered taxa, or (2) the increased presence of potentially predator taxa such as ants in natura results in a regulation of these parasite populations, as suggested by Lambrechts et al [26]. The first possibility is not unlikely to occur at least with O. sylviarum as it has already shown being able to develop resistance to inhibitors of acetylcholinesterase in the USA [30]. Indeed, populations in this species are regularly exposed to organophosphates (Ravap EC) or carbamates (sevin-80s) in North American poultry farms, not only by spraying, but also in some cases by digging [31, Rubinoff I, pers. comm.]. And yet, it is also exposed to various pesticides in French pet birds breeding facilities, including carbamates and organophosphates [Association Ornithologique Rhodanienne, pers. comm.], which may have resulted in a selection of resistant populations in France too. This may explain its presence in the three control conditions under test (plus in natura). As for D. gallinae, it is regularly exposed to organophosphates and other pesticides (such as amitraze, for instance) in European farms (legally during the empty period, but also illegally during flocks), but the hypothesis of selected populations here is not so consistent, as it was found only in integrated and organic orchards, along with the unselected feral D. carpathicus species. Anyway, the presence of D. gallinae in orchards and its absence in natura nests may be due to concomitant enhanced tolerance to neutral pesticide substances (as encountered in organic farms also, such as sulphur) and reduction of predator presence (second possibility). Lesna et al [29] showed significant negative relationships between the mite predator Androlaelaps casalis and D. gallinae densities in some starling nest boxes. But in any case, populations of Dermanyssus in tit nests in natura are restricted to D. carpathicus and D. longipes, whereas in orchards, D. gallinae is also regularly found. Here is an illustration of results of Ives et al [32], who demonstrated that species richness increases community-level stability by insuring that some species in a community are tolerant to different environmental fluctuations. Among recurrent arthropoda in bird nests, the occurrence of a less species-rich group (detritivorous / saprophagous mites – one largely dominant species: H. chelidonis) is more affected by agricultural practices in general, than are more species-rich nidicolous parasites. And yet, two of these species (D. gallinae and O. sylviarum) are known tolerating different environmental conditions than the three others. On the other hand, the psocopteran order is composed of almost a single species, L. bostrychophila (the second psocopteran species having been found only once and being known to live under tree bark) and is not as affected by agricultural practices as do detritivorous Prostigmata. But L. bostrychophila is a parthenogenetic species and possesses an amazing flexibility, which allows it rapidly getting more tolerant to a variety of conditions. It can easily become resistant to organophosphates and carbamate [33, 34] and proved to adapt to hypoxia and hypercarbia [35]. It showed even some adaptation against entomopathogenic fungi [36] and is not so sensitive to a bacterium-derived insecticide, which is efficient against other insect pests [37]. Cross resistance between very different parameters (hypoxia, hypercarbia, organophosphates,…) have been evidenced by Wei et al [38], so that the last mean found to control this species is the use of an antibiotic in order to kill an endosymbiotic bacteria and so slightly reduce the pest's fitness [39]. The occurrence of this species indifferently in all orchard conditions might be due to its extreme flexibility, allowing it to adapt, alone, to all conditions under test.
224 Publication VI
4.2.3. Hematophagous Mesostigmata The five species of hematophagous Mesostigmata recorded in the present study belong to two distant families within Dermanyssoidea (Dermanyssidae and Macronyssidae). They correspond to species encountered in natura on similar hosts [19, 20]. Of them, only O. sylviarum was noted in conventional orchards, and it was repeatedly found in such conditions. This species is a pest in layer farms in USA [40, 41]. D. gallinae, the equivalent pest in European layer farms [42, 43], has only been encountered in organic or integrated orchards. Note that O. sylviarum is currently a pest in North American layer farms, but not in French layer farms, contrarily to what was published by Bruneau et al [44]. Roy et al [19, 20, 45] sampled mites the same way in same regions in more than 40 layer farms and always found D. gallinae alone. Not only present mt-Co1 phylogenetic analysis show that the American isolate OSBM, the African isolate [21] and isolates in present study are conspecific (Fig. 5), but also 16S RNA sequences suggest that French field isolates are not more distant from American than from African isolates (acc. no AY185362). This is in favour of a close relationship between population from layer farms (American isolate OSBM) and from wild avifauna (French and African samples). A similarly close relationship between poultry and wild populations of D. gallinae is visible in [19] (isolates D. gallinae JBOn). At least in farms, these two species do not have same habits. O. sylviarum not only stays long on host, but also lays on it, so that a direct examination of the bird allows detecting a large number of mites in farms [41, LR pers. observation]. In contrast, D. gallinae has exclusively nidicolous habits, has a fast blood meal, about as fast as do mosquitoes or more appropriately as do bed bugs, and once the meal is completed, the mite quickly goes back to its hiding-place [46]. Eggs are laid in hiding-places, never on bird [47]. Does this result in less sensitive O. sylviarum than D. gallinae to some pesticides, since D. gallinae is more difficult to reach with sprayed pesticides? Such an impact could explain the difference noted in present study between these two taxa. Anyway, this difference is not well supported by statistical tests, apparently due to the too weak number of samples containing Mesostigmata (and more generally samples containing any arthropoda). Nevertheless, O. sylviarum was also detected in abundance in wild avifauna by authors in three nests of Circus pygargus (Montagu's Harrier) sampled in wheat meadows, which are chemically controlled and where no D. gallinae was found. On the other hand, never D. gallinae has been noted by authors before in great tits' nests (Parus major) [19]. It has been isolated in present study in 2007 as well as in 2008 orchard nests. This species has been shown to be much more synanthropic than tougher D. carpathicus and D. longipes and is likely to resist less to predators' activity [19, 20]. And yet, ants seem to be less present in orchards nests than in natura. Even organic control methods may have an impact on such auxiliaries arthropoda, whereas D. gallinae and O. sylviarum seem to be more tolerant, especially the latter. 5. Conclusion In conclusion, the presence/absence of any arthropod in bird nests is a reliable indicator of some agricultural practices' impact in orchards. A marked contrast between the two terminal agroecosystem modalities and an intermediate characterisation of integrated control orchards were observed using the presence/absence of arthropoda approach. The diversity within arthropodan communities does not provide as sharp information as did the simple presence / absence notation. Nevertheless, two main elements seem to arise from our diversity data set: (1) among hematophagous mesostigmata species, not all seem to tolerate chemical methods in orchard nests, (2) the diversity of arthropoda communities is likely to be more important in natura nests. 225 Publication VI
Finally, it is remarkable that O. sylviarum is so commonly encountered in bird nests under test, yet absent from French layer farms, whereas it is a serious pest in layer farms in the USA. The opposite applies to D. gallinae, the layer pest in Europe. This suggests that different farming practices between both continents might explain the paradox of the omnipresence of both species in wild avifauna and a selective presence in layer farms. 6. Acknowledgments Authors want to warmly thank Z. Kucerova (Crop Research Institute, Prague, Czech Republic), R. Allemand (Université Lyon 1, Villeurbanne, France) and F. Faraji (MITOX Consultants, Amsterdam, The Netherlands) for insect or mite identifications, M. Sabelis and I. lesna (IBED, University of Amsterdam, Amsterdam, The Netherlands) for valuable comments, B. Mullens (University of California, Riverside, USA), F. Salmon (Centre de Recherche sur la Biologie et les populations d’Oiseaux, Paris, France), A. Marchal and ? (Association Ornithologique Rhodanienne, Lyon, France), S. Calabro (Association Ornithologique Becs Crochus Centre Est, Relevant, France), I. Rubinoff (Hy-Line International, Dallas Center, USA) for useful samples and information. Finally, LR would like to offer her most sincere thanks for his technical participation to G. Lallemand (LEGTA des Mandailles, Chateauneuf-de-Galaure, France). 7. References [1] P. Duelli, M.K. Obrist, D.R. Schmatz, Biodiversity evaluation in agricultural landscapes: above-ground insects, Agriculture, Agric. ecosyst. environ. 74 (1999) 33-64. [2] N.C. Elliot, R.W. Kieckhefer, J.H. Lee, B.W. French, Influence of within-field and landscape factors on aphid predator populations in wheat, Landsc. ecol. 14 (1998) 239-252. [3] T. Purtauf, I. Roschewitz, J. Dauber, C. Thies, T. Tscharntke, V. Wolters, Landscape context of organic and conventional farms : influences on carabid beetle diversity, Agric. ecosyst. environ. 108 (2005) 165-174. [4] A.C. Weibull, J. Bengtsson, E. Nohlgren, Species composition in agroecosystems : the effect of landscape, habitat, and farm management, Basic appl. ecol. 4 (2003) 349-361. [5] A. MacLeod, S. Wratten, N. Sotherton, M. B. Thomas, 'Beetle banks' as refuges for beneficial arthropods in farmland: long-term changes in predator communities and habitat, Agric. for. entomol. 6 (2004) 147-154. [6] J.C.Bouvier, J.F. Toubon, T. Boivin, B. Sauphanor, Effects of apple orchard management strategies on the great tit (Parus major) in south-eastern France, Envir. Toxicol. Chem. 24 (2005) 2846-2852. [7] E. Padoa-Schioppa, M.Baietto, R. Massa, L. Bottoni, Bird communities as bioindicators: The focal species concept in agricultural landscapes, Ecological indicators, 6 (2006) 83-93. [8] J. Cabral, A. Rocha, M. Santos, A. Crespi, A stochastic dynamic methodology to facilitate handling simple passerine indicators in the scope of the agri-environmental measures, Ecological indicators, 17 (2007) 34-37. [9] P. Zeman, M. Jurík, A contribution to the knowledge of fauna and ecology of gamasoid mites in cavity nests of birds in Czechoslovakia, Folia Parasitologia 28 (1981) 265-271. [10] E. H. Burtt Jr., W. Chow, G.A. Babbitt, Occurrence and demography of mites of Tree Swallow, House Wren, and Eastern Bluebird nests, in : J. E. Loye and M. Zuk, editors. Bird- parasite interactions: ecology, evolution, and behaviour, Oxford University Press, Oxford, UK, 1991, pp. 104-122. [11] P. Fenća, E. Schniererová, Mites (Acarina: Mesostigmata) in the nest of Acrocephalus spp. and in neighbouring reeds, Biologia, 59 (2004) 41-47.
226 Publication VI
[12] J. Nosek, M. Lichard, Beitrag zur Kenntnis der Vogelnestfauna, Entomol. probl. 2 (1962) 29-51. [13] A. Fain, T.D. Galloway, Mites (Acari) from nests of sea birds in New Zealand. Ii. Mesostigmata and Astigmata, Bull. - Inst. r. sci. nat. Belg., Entomol. 63 (1993) 95-111. [14] C.G. Majka, J. Klimaszewski, R.F. Lauff, New Coleoptera records from owl nests in Nova Scotia, Canada, Zootaxa 1194 (2006) 33-47. [15] O.Merkl, J. Bagyura, L.Rózsa, Insects inhabiting Saker (Falco cherrug) nests in Hungary, Ornis Hungarica 14 (2004) 1-4. [16] N. Galtier, M. Gouy, C. Gautier, SEAVIEW and PHYLO_WIN: two graphic tools for sequence alignment and molecular phylogeny, CABIOS 12 (1996) 543-548. [17] Z. Šustek, J. Kridžtofik, Beetles (Coleoptera) in deserted nests of Phoenicurus ochruros, Parus caeruleus, Parus major, Sitta europaea and Sturnus vulgaris, Entom. carpath. 14 (2002) 64-69. [18] L. Roy, C.M. Chauve, G. Lallemand, T. Buronfosse, Implication du genre Dermanyssus dans l’arthropodofaune des nids d’oiseaux sauvages, Congrès de la Société Française de Parasitologie, Nice, 2007, 13-14th december. [19] L. Roy, A.P.G. Dowling, C.M. Chauve, I. Lesna, M.W. Sabelis, T. Buronfosse, Molecular phylogenetic assessment of host range in five Dermanyssus species, Exp. Appl. Acarol. 48 (2009) 115-142. [20] L. Roy, A.P.G. Dowling, C.M. Chauve, T. Buronfosse, Delimiting species boundaries within Dermanyssus Dugès, 1834 (Acari:Dermanyssidae) using a total evidence approach, Mol. Phylogenet. Evol. 50 (2009) 446-470. [21] M.D. Schrenzel, G.A. Maalouf, L.L. Keener and P.M. Gaffney, Molecular characterization of malarial parasites in captive passerine birds, j. parasitol. 89 (2003) 1025- 1033. [22] I. Roschewitz, M. Hücker, T. Tscharntke, T. Carsten, The influence of landscape context and farming practices on parasitism of cereal aphids, Agric. ecosyst. environ. 108 (2005) 218- 227. [23] J. Krištofík, Z. Šustek, P. Mašán, Arthropods (Pseudoscorpionida, Acari, Coleoptera, Siphonaptera) in the nest of red-backed shrike (Lanius collurio) and lesser grey shrike (Lanius minor), Biologia 57 (2002) 603-613. [24] J. Krištofík, P. Mašán, Z. Šustek, Arthropods (Pseudoscorpionidea, Acarina, Coleoptera, Siphonaptera) in nests of the bearded tit (Panurus biarmicus), Biologia 62 (2007) 749-755. [25] D.C. Duffy. Ants, ticks, and nesting seabirds: dynamic interactions, in: J. E. Loye and M. Zuk, editors, Bird-parasite interactions: ecology, evolution, and behaviour, Oxford University Press, Oxford, UK, 1991, pp. 242-257. [26] M.M. Lambrechts, B. Schatz, P. Bourgault, Interactions between ants and breeding Paridae in two distinct Corsican oak habitats, Folia Zool. 57 (2008) 264-268. [27] Z. Šustek, D. Hornychová, The Beetles (Coleoptera) in the nests of Delichon urbica in Slovakia, Acta Rer. natur. Mus. nat. slov. 29 (1983) 119-134. [28] C. Lienhard, Psocoptères euro-méditerranéens, Faune de France, Fédération Française des Sociétés de Sciences naturelles, 1998, 83. [29] I. Lesna, P. Wolfs, F. Faraji, L. Roy, J. Komdeur, M.W. Sabelis, Candidate predators for biological control of the poultry red mite Dermanyssus gallinae, Exp. Appl. Acarol. 48 (2009) 63-80. [30] B.A. Mullens, R.K. Velten, N.C. Hinkle, D.R. Kuney, C.E. Szijj, Acaricide resistance in northern fowl mite (Ornithonyssus sylviarum) populations on caged layer operations in Southern California. Poult. sci. 83 (2004) 365-374. [31] B.A. Mullens, J.P. Owen, D.R. Kuney, C.E. Szijj, K.A. Klingler, Temporal changes in distribution, prevalence and intensity of northern fowl mite (Ornithonyssus sylviarum)
227 Publication VI parasitism in commercial caged laying hens, with a comprehensive economic analysis of parasite impact, Veterinary Parasitology 160 (2009) 116-133. [32] A.R. Ives, J.L. Klug, K. Gross, Stability and species richness in complex communities, Ecol. lett. 3 (2000) 399-411. [33] M.K. Nayak, P.J. Collins, R.A. Kopittke, Residual toxicities and persistence of organophosphorus insecticides mixed with carbaryl as structural treatments against three liposcelidid psocid species (Psocoptera: Liposcelididae) infesting stored grain, J. Stored Prod. Res. 39 (2003) 343-353. [34] Chaia Yu-Xin, Liua Guo-Ying, Wang Jin-Jun, Toxicological and biochemical characterizations of AChE in Liposcelis bostrychophila Badonnel (Psocoptera: Liposcelididae), Pestic. biochem. physiol. 88 (2007) 197-202. [35] Wang Jinjun, Zhao Zhimo, Li Lungshu, Some Biochemical Aspects of resistance to controlled atmosphere in Liposcelis bostrychophila Badonnel (Psocoptera: Liposcelididae). Insect sci. 6 (2009) 178-186. [36] J.C.Lord, R.W. Howard, A proposed role for the cuticular fatty amides of Liposcelis bostrychophila (Psocoptera: Liposcelidae) in preventing adhesion of dry-conidia entomopathogenic fungi. Mycopathologia 158 (2004) 211-217. [37] M.K. Nayak, G.J. Daglish, V.S. Byrne, Effectiveness of spinosad as a grain protectant against resistant beetle and psocid pests of stored grain in Australia, J. Stored Prod. Res. 41 (2005) 455-467. [38] Wei Ding, ZhiMo Zhao, JinJun Wang, HuiYing Tao, YongQiang Zhang, The relationship between resistance to controlled atmosphere and insecticides of Liposcelis bostrychophila Badonnel (Psocoptera: Liposcelididae), Agric. sci. China [39] Wang Jinjun, Dong Peng, Xiao Li-Sha, and Dou Wei, Effects of Removal of Cardinium Infection on Fitness of the Stored-Product Pest Liposcelis bostrychophila (Psocoptera: Liposcelididae), J. econ. entomol. 101(2008):1711-1717. [40] R.C. Axtell, J.J. Arends, Ecology and Management of arthropod pests of poultry, Ann. Rev. Entomol. 35 (1990) 101-126. [41] B.A. Mullens, N.C. Hinkle, L. J. Robinson, C.E. Szijj, Dispersal of Northern Fowl Mites, Ornithonyssus sylviarum, Among Hens in an Experimental Poultry House, J. appl. poult. res. 10 (2001) 60-64. [42] JH Guy, M Khajavi, MM Hlalel, O. Sparagano, Red mite (Dermanyssus gallinae) prevalence in laying units in Northern England, Br. poult. sci. 45 (2004) 15-16. [43] O. Sparagano, A. Pavliüeviü, T. Murano, A. Camarda, H. Sahibi, O. Kilpinen, M. Mul, R van Emous, S. le Bouquin, K. Hoel, M. Cafiero, Prevalence and key figures for the poultry red mite. Dermanyssus gallinae infections in poultry farm systems, Exp. Appl. Acarol. 48 (2009) 3-10. [44] A. Bruneau, A. Dernburg, C. Chauve, L. Zenner, First report of the northern fowl mite Ornithonyssus sylviarum in France, Vet. record. 150 (2002) 413-414. [45] L. Roy, C.M. Chauve, J. Delaporte, G. Inizan, T. Buronfosse, Exploration of the susceptibility of AChE from the Poultry Red Mite Dermanyssus gallinae (Acari: Mesostigmata) to organophosphates in field isolates from France, Exp. Appl. Acarol. 48 (2009) 19-30. [46] H.P. Wood, The chicken mite : its life history and habits. United States Department of Agriculture, Washington, DC., Bull. 553 (1917) 1-14. [47] W.W. Moss, The mite genus Dermanyssus: a survey, with description of Dermanyssus trochilinis, n. sp., and a revised key to the species (Acari: Mesostigmata: Dermanyssidae), J Med Ent 14 (1978) 627-640.
228 Publication VI
8. Legends Tables Table 1. Species detected in the four focused recurrent primary groups. ? represent larvae potentially belonging to corresponding species. * Rare, ** Common, yet never abundant, *** common and sometimes abundant
Table 2. Number of species within Coleoptera and Mesostigmata in arthropod-rich and arthropod-poor nests sampled in 2007.
Figure Figure 1. Distribution of arthropod-rich, arthropod-poor and arthropod-free nests according to control conditions. Three types of orchard: AB = organic, Int. = integrated and Chem. = conventional. Nests from natura have been sampled in not human-shaped environments.
Figure 2. Percentage of occurrence of arthropoda groups as defined in material and methods. A black star indicates a group identified at the specific level.
Figure 3. Average number of arthropodan groups in arthropod-rich or arthropod-poor nests
Figure 4. Number of nests containing one or several species belonging to the 2 hematophagous mite families isolated in present study according to control management. No nest from nature has been sampled in 2008.
Figure 5. Phylogenetic topology involving individuals of D. gallinae and O. sylviarum isolated in the framework of present study (sampled JBOn), along with some individuals obtained from different environments and places. Maximum Likelihood, Phylo_win. Numbers at nodes represent bootstrap values for 500 replicates.
229 Publication VI
9. Appendix 1. Host and location information for tit nests sampled in natura.
French Nest Bird species departement Parus LC 7 caeruleus Drôme LBO7-1 Parus major Loiret LBO7-2 Parus major Loiret LC10 Parus major Drôme LC4 Parus major Drôme LC6 Parus major Drôme MA1 Parus major Rhône MA2 Parus major Rhône N4 Parus major Rhône N5 Parus major Rhône N6 Parus major Rhône N7 Parus major Rhône OC2 Parus major Rhône JGC3 Parus sp. Drôme JGC4 Parus sp. Drôme JGC5 Parus sp. Drôme N1 Parus sp. Rhône N2 Parus sp. Rhône N3 Parus sp. Rhône N9 Parus sp. Rhône NO18 Parus sp. Loire
230 Publication VI
10. Appendix 2. Mite species, sample location and EMBL accession numbers for sequenced samples
16S COI rRNA Dermanyssus French accession accession species Host Country department Context Environment Individual numbers numbers bird nest - inside Colony inside a building, in a a building, in a D. apodis Swift France Gard town town GO54c id. AM921874 D. bird nest - nest carpathicus Tit France Rhône box Nature reserve Ecop3a FM208729 D. Bouches- bird nest - nest Organic carpathicus Tit France du-Rhône box orchard JBO133f xxxxxxxx D. Bouches- bird nest - nest Organic carpathicus Tit France du-Rhône box orchard JBO134d xxxxxxxx D. Bouches- bird nest - nest Organic carpathicus Tit France du-Rhône box orchard JBO134f xxxxxxxx D. Bouches- bird nest - nest Organic carpathicus Tit France du-Rhône box orchard JBO135a D. Bouches- bird nest - nest Organic carpathicus Tit France du-Rhône box orchard JBO135B xxxxxxxx D. Bouches- bird nest - nest Organic carpathicus Tit France du-Rhône box orchard JBO135f xxxxxxxx D. Bouches- bird nest - nest integrated carpathicus Tit France du-Rhône box orchard JBO83b xxxxxxxx D. Pas-de- bird nest - nest Individual carpathicus Tit France Calais box garden JMC10A AM943021
231 Publication VI
Individual car D. bird nest - girder park in the carpathicus Redstart France Loire within a building country RQ20 xxxxxxxx Individual car D. bird nest - girder park in the carpathicus Redstart France Loire within a building country RQ24 xxxxxxxx conventional D. gallinae Layer hen France Ain layer farm layer farm 8006b xxxxxxxx conventional D. gallinae Layer hen France Ain layer farm layer farm 8006B16c xxxxxxxx Amateur pigeon pigeon breeding breeding D. gallinae Pigeon France Rhône facility - aviary facility 8008g FM208712 chicken farm - AOC chicken D. gallinae Layer hen France Ain Bresse (epinettes) farm 8012a FM208739 Bouches- bird nest - nest Organic D. gallinae Tit France du-Rhône box orchard JBO461 FM208736 Bouches- bird nest - nest Organic id. D. gallinae Tit France du-Rhône box orchard JBO464 FM208736 Bouches- bird nest - nest Organic D. gallinae Tit France du-Rhône box orchard JBO75a xxxxxxxx Bouches- bird nest - nest Organic D. gallinae Tit France du-Rhône box orchard JBO75b xxxxxxxx Bouches- bird nest - nest Organic D. gallinae Tit France du-Rhône box orchard JBO75c xxxxxxxx Bouches- bird nest - nest integrated D. gallinae Tit France du-Rhône box orchard JBO90a xxxxxxxx Bouches- bird nest - nest integrated D. gallinae Tit France du-Rhône box orchard JBO90b xxxxxxxx Bouches- bird nest - nest integrated D. gallinae Tit France du-Rhône box orchard JBO90c xxxxxxxx
232 Publication VI
bird nest - from AM921867 D. gallinae House martin France Ain8 barn Barn LB183 conventional D. gallinae Layer hen Poland layer farm layer farm PO2A xxxxxxxx Bouches- D. gallinae Roller France du-Rhône on bird - nest box Natura ROL23 AM921865 Bouches- D. gallinae Roller France du-Rhône on bird - nest box Natura ROL26 xxxxxxxx lab strain Lab strain (original isolate sampled in a from a layer conventional D. gallinae Layer hen Denmark farm) layer farm SK0815 xxxxxxxx lab strain Lab strain (original isolate sampled in a from a layer conventional D. gallinae Layer hen Denmark farm) layer farm SK0818 xxxxxxxx on bird - banding D. gallinae Pic France Rhône activity Natura Woodp AM921863 bird nest - nest D. hirundinis Tit USA box BMOC FM208746 bird nest - nest D. longipes Tit France Rhône box Park ENVL083a FM179365 bird nest - nest D. longipes Tit France Rhône box Park ENVL088a FM208743 bird nest - nest D. longipes Tit France Rhône box Park PM xxxxxxxx Bouches- bird nest - nest Organic D. longipes Sparrow France du-Rhône box orchard JBO108a xxxxxxxx Bouches- bird nest - nest Organic D. longipes Sparrow France du-Rhône box orchard JBO495 AM921869 Bouches- bird nest - nest conventional O. sylviarum Tit France du-Rhône box orchard JBO78c xxxxxxxx
233 Publication VI
Bouches- bird nest - nest Organic O. sylviarum Tit France du-Rhône box orchard JBO74a xxxxxxxx Bouches- bird nest - nest Organic O. sylviarum Tit France du-Rhône box orchard JBO100b xxxxxxxx bird nest - nest Organic O. sylviarum Tit France Ain4 box orchard JBO122 xxxxxxxx Amateur pet bird breeding O. sylviarum Canary France Isère cage - litter facility Rh1b xxxxxxxx Amateur pet bird breeding O. sylviarum Canary France Isère cage - litter facility Rh1d xxxxxxxx Montagu's bird nest - conventional O. sylviarum Harrier France Vendée wheatfield wheatfield FS5a xxxxxxxx Montagu's bird nest - conventional O. sylviarum Harrier France Vendée wheatfield wheatfield FS5b xxxxxxxx Montagu's bird nest - conventional O. sylviarum Harrier France Vendée wheatfield wheatfield FS5c xxxxxxxx Montagu's bird nest - conventional O. sylviarum Harrier France Vendée wheatfield wheatfield FS6a xxxxxxxx Montagu's bird nest - conventional O. sylviarum Harrier France Vendée wheatfield wheatfield FS6a xxxxxxxx conventional O. sylviarum Layer hen USA layer farm layer farm OSBM xxxxxxxx xxxxxxxx outgroup - lab O. bacoti Rodent France strain OBAC xxxxxxxx outgroup - lab T. pyri France Hérault strain T_pyri FM179364 outgroup - bird A. casalis France Rhône nest ACA AM921868 xxxxxxxx
234 Publication VI
Tables
Table 1.
Primary group Species / family Environment % occurrence - * abundance within nest Organic Integrated Convention natura orchards orchards al orchards Acari hematophagous D. gallinae (Dermanyssidae) 6,6 ** 3,6 ** 0,0 0,0 - Mesostigmata D. carpathicus (Dermanyssidae) 11,5 ** 5,4 ** 0,0 9,5 ** D. longipes (Dermanyssidae) 6,6 ** 0,0 0,0 9,5 ** Ornithonyssus sylviarum (Macronyssidae) 11,5 ** 7,1 ** 14,6 ** 4,8 ** Ornithonyssus sp (Macronyssidae) 1,6 * 1,8 * 0,0 0,0 - Remark : this species is very similar to, but slightly differs from O. bacoti (Hirst, 1913), according to Micherdziƌski (1980) and compared to individuals of O. bacoti of a lab strain in MNHN detritivorous Hirstia chelidonis (Astigmata : 30,8 *** 21,1 *** 2,1 ** Acariforms Pyroglyphidae) Tyrophagus longior (Astigmata : Acaridae) 0,0 2,6 * 4,2 * Insecta Psocoptera Liposcelis bostrychophila (Liposcelididae) 59,0 *** 31,6 ** 11,4 * Embidopsocus enderleini (Liposcelididae) 0,0 2,6 * 0,0 Coleoptera Attagenus unicolor (Dermestidae) ? ? ? 4,8 * Anthrenus pimpinellae (Dermestidae) ? ? ? 9,5 *** Dermestes undulatus (Dermestidae) ? ? ? 4,8 ** Cryptophagus sp (Cryptophagidae) 0,0 0,0 0,0 4,8 * Alphitobius diaperinus (Tenebrionidae) 0,0 0,0 0,0 4,8 * Nalassus dryadophilus (Tenebrionidae) 0,0 0,0 0,0 4,8 * Gnathoncus buyssoni (Histeridae) 0,0 0,0 0,0 4,8 * Carcinops 14 striatus (= pumilio) 2,6 * 0,0 0,0 4,8 * (Histeridae) Potosia oblonga (Scarabaeidae) 0,0 2,6 * 0,0 0,0 -
Table 2.
No of species (n=no of Average number of species nests under test) per nest in natura orchards in natura orchards (n=19) (n=43) Coleoptera 7 2 1.6 1.0 Mesostigmata 3 5 1.0 1.1
235 100% 2007
80%
60%
40%
20%
0% wild avifauna AB orchard (n=27) int. Orchard (n=21) chem. Orchard (n=21) (n=14) 100% 2008 80%
Arthropod-rich nests 60% Arthropod-poor nests Arthropod-free nests 40%
20%
0% AB orchard (n=39) int. orchard (n=38) chem. orchard (n=35) Figure 1
Nests sampled in natura 2007 (n=21)
60
50
40
30
20 % occurence 10
0
Orchard nests 2007 (n=63)
60 50
40
30 20 % occurence % 10 0 Orchard nests 2008 (n=110)
60
50
40
30
20 % occurence 10
0 a ) ta t a a ) a a a a ) la a s a a id e s) s) m era ter er o er ic p r did ne t t t pt l ea ri bo fo o ra ap fl h m opod tigma i Ix Dipter t s ar A m nop ing I ostigma c r e oleopteremi w a( a ( olle s sos socop e C H r europteraC e e A P D (parasi Lepidopter e N M M ym pter us (H era opt us us o a t ona o o vor de iph riti S hysan et coi T D mi hymenop Figure 2 or o Mallophaga (she F ematophag cr H Mi
Non hematophagCheleytoidea (predator Acariforms)
1 2007 4,0 3,5
3,0
2,5
2,0
1,5
1,0
0,5
0,0 natura AB int. chem.
4 2008 3,5
3
2,5
2
1,5
1
0,5
0 AB int. chem. Figure 3
90 80 70 60 Chem. 50 Int. 40 AB natura 30 20 10 0 Dermanyssus 2007 Dermanyssus 2008 Ornithonyssus 2007 Ornithonyssus 2008
Figure 4
2 O. bacoti
O. sylviarum
D. apodis D. hirundinis D. longipes
D. carpathicus
D. gallinae
Figure 5
3
6 Discussion
6.1 Relations phylogénétiques
a - Des lacunes Le genre Liponyssoides demeure un mystère, puisque nous n’avons pu obtenir aucun type et n’avons jamais pu en isoler un seul individu sur le terrain (cf. § 4.3.b.3 p. 67). La famille des Dermanyssidae n’englobant que Dermanyssus et Liponyssoides (cf. Annexe 1), il en résulte que seuls des outgroups distants ont pu être intégrés aux analyses phylogénétiques (familles des Macronyssidae, Laelapidae, Phytoseiidae ; cf. Annexe 1). Certes, ces outgroups permettent un enracinement sans ambiguïté des populations testées du genre Dermanyssus. Mais la monophylie du genre Dermanyssus vis-à-vis de son plus proche parent demeure à tester. Il n’est pas certain que la séparation entre les espèces des deux genres soit une séparation naturelle. Au sein de Dermanyssus, seul le sous-genre Dermanyssus a pu être testé par l’approche « exhaustive » (morphologie + marqueurs moléculaires), dont seulement 2 espèces du groupe hirsutus. Il apparaît toutefois assez clair que les subdivisions dans le sous-genre Dermanyssus ne correspondent pas à deux groupes monophylétiques, comme le considérait Moss. D. hirsutus et D. quintus s’embranchent au beau milieu des espèces du groupe gallinae sur la base des quatre marqueurs moléculaires développés. En outre, le sous-genre Microdermanyssus, sur la base de l’analyse morphologique de l’ensemble du genre (Publication III), semble apparenté à D. hirsutus et D. quintus. Cela suggère aussi l’invalidité de la subdivision en deux sous-genres, mais demeure à vérifier.
b - Topologies bifides ou en escalier ? D’une manière générale, les relations distales des populations testées au long de la présente étude sont bien résolues sur la base de chacun des deux gènes mitochondriaux, présentent des supports importants suivant les deux critères utilisés (maximum de parcimonie - MP, méthode bayésienne -BA), tandis que les relations plus basales (internes) demeurent irrésolues sur la base de l’ARNr 16S et au moins en partie faiblement soutenues sur la base de la mt-Co1. Les entités spécifiques apparaissent ainsi bien délimitées, mais leurs relations mutuelles ne sont pas claires sur la base mitochondriale. Cela s’accorde bien avec la transmission clonale du génome mitochondrial, résultant en une taille efficace de la population Ne réduite au tiers comparé au génome nucléaire. La vitesse d'évolution des gènes mitochondriaux est par conséquent plus importante. Les marqueurs nucléaires se comportent très différemment des marqueurs mitochondriaux, mais aussi entre eux. Les reconstructions phylogénétiques obtenues sur la base des ITS1 et 2 demeurent fortement irrésolues, tant distalement que basalement (Publication III). Toutefois, une structure en escalier apparaît sur cette base, même si faiblement soutenue. Sur la base de l’EF1-alpha, une anomalie de type paralogie rend le marqueur inutilisable en l’état, puisque des outgroups se voient intégrés au beau milieu de Dermanyssus (cf. § 4.3.b.2). Enfin, l’amplicon de Tropomyosine, intégrant deux courtes portions d’exon et une portion plus conséquente d’intron offre une résolution importante aussi bien dans les relations internes que distales (Publication V). Les entités spécifiques clairement délimitées sur la base mitochondriale correspondent toutes aux clades monospécifiques des topologies obtenues sur la base de la Tropomyosine, exceptée D. longipes qui se révèle polyphylétique (cf. plus bas). Les relations basales aussi sont bien résolues, les indices de cohérence (CI) et de rétention (RI) obtenus en MP indiquent un taux d’homoplasie plutôt bas.
239 Deux clades se dessinent dans les analyses phylogénétiques basées sur la mt-Co1 (Publication IV et V ; cf. Fig. 6), regroupant pour l’un D. hirundinis, D. longipes et D. carpathicus (clade A ou groupe hirundinis), pour l’autre les diverses lignées de D. gallinae (clade B, publications III, IV, V). En revanche, sur la base de la Tropomyosine (publication V), la topologie retenue, ie celle obtenue en MP, avec les gaps considérés comme un 5ème état, présente les relations interspécifiques entre les 5 espèces testées et avec D. hirsutus comme graduelles plutôt que bifides (Publication V ; cf. Fig. 7): le clade A observé sur la base mitochondriale apparaît paraphylétique. Mais l’ordre de leur branchement mutuel n’est pas remis en cause. D. hirundinis et les deux lignées de D. longipes occupent les positions les plus basales. D. carpathicus représente ensuite le groupe frère de l’ensemble des restants, suivi de D. hirsutus. Enfin, D. apodis et les différentes lignées de D. gallinae constituent les 2 entités sœurs en position distale dans la topologie. Les relations de D. apodis avec l’ensemble des autres espèces demeuraient incertaines dans les topologies mitochondriales (supports de nœuds faibles), quoique dans tous les cas à la base de l’un ou l’autre des clades principaux (A ou B, cf. Publication IV). Elles s’avèrent finalement plutôt en faveur d’une relation étroite avec les lignées de D. gallinae sur la base du dernier marqueur nucléaire développé (Tropomyosine, Publication V). Non seulement les supports pour cette topologie sont satisfaisants, mais encore de nombreuses études font état d’une faiblesse des marqueurs mitochondriaux dans la résolution des relations internes. Cela nous pousse à choisir les topologies basées sur la Tropomyosine en ce qui concerne les relations de D. apodis avec les autres espèces. Dans tous les cas, D. apodis se présente à l’interface entre représentants du clade A et du clade B mitochondriaux et possède des caractères écologiques que l’on peut considérer comme intermédiaires entre les deux clades (cf. ci-dessous § Patterns écologiques révélés). Enfin, la position interne de D. hirsutus, seul représentant du groupe hirsutus de Moss, dans nos topologies basées sur la Tropomyosine (publication V) vient confirmer l’invalidité de la division opposant groupe gallinae à groupe hirsutus, déjà pressentie sur la base des ITS (publication III). Les deux marqueurs retenus pour l’exploration intraspécifique sont aussi les marqueurs qui offrent la meilleure résolution des relations à tous les niveaux. 6.2 Etat de la taxinomie du genre Dermanyssus Le groupe gallinae sensu Moss s’est révélé beaucoup plus riche en espèces que nous ne l’avions supposé au début de l’étude. En effet, D. hirundinis et D. longipes, qui a priori paraissaient indiscernables de D. gallinae, se sont révélés clairement isolés des autres espèces, dont D. gallinae, sur le plan reproducteur. D. longipes, longtemps considéré incertae sedis, puis transféré nomen dubium dans la publication I, a finalement affirmé son existence, à la suite de prélèvements dans la région type, dans la publication III. En outre, une seconde espèce, indiscernable morphologiquement de D. hirundinis et de D. longipes, a été considéré comme appartenant à cette dernière espèce dans la publication III et déclarée espèce cryptique à décrire dans la publication V (lignée ENV). D. carpathicus a confirmé aussi son statut spécifique, comme pouvait le laisser attendre les quelques caractères discriminants notés a priori. Une espèce nouvelle a été décrite : D. apodis. Enfin, D. gallinae s’avère constitué d’une diversité de lignées, dont l’une semble être, en fait, une espèce cryptique, divergeant encore faiblement des autres (la lignée L1). Enfin, les subdivisions de Moss ont été remises en cause (cf. Fig. 8).
240 Genre Espèces
Dermanyssus Dugès, 1834 D. antillarum Dusbabek and Cerny, 1971 D. chelidonis Oudemans, 1939 D. faralloni Nelson and Furman, 1967 D. gallinoides Moss 1966 Interrelations et relations avec les D. nipponensis Uchikawa and Kitaoka, 1981 autres entités non résolues sur la D. prognephilus Ewing, 1933 base morphologique D. transvaalensis Evans and Till, 1962 D. triscutatus Krantz, 1959 (cf. Publication III) D. trochilinis Moss, 1978 D. wutaiensis Gu and Ting, 1992 *D. carpathicus Zeman, 1979 *D. hirundinis (Hermann, 1804) *D. longipes Berlese and Trouessart, 1889 (= PAS ; cf. Publication V) *D. longipes ENVL08 espèce cryptique probable *D. gallinae De Geer, 1778 s. stricto complexe de lignées/espèces naissantes *D. gallinae L1 espèce cryptique probable Clade D *D. apodis Roy, Dowling, Chauve & Buronfosse, 2009 (cf. Publication V) D. hirsutus Moss and Radovsky 1967 D. alaudae (Schrank, 1781) D. americanus Ewing 1922 D. brevirivulus Gu and Ting, 1992 Groupe Microdermanyssus + D. brevis Ewing, 1936 groupe hirsutus D. grochovskae Zemskaya 1961 D. passerinus Berlese and Trouessart, 1889 (species inquirenda) > Clade isolé sur la base D. quintus Vitzthum, 1921 morphologique (cf. Publication D. rwandae Fain, 1993 III) > Embranchement probable à la base du clade D sur la base moléculaire nucléaire (cf. D. diphyes Knee, 2008 Espèce à la morphologie Publication V) intermédiaire entre clade Microdermanyssus + groupe hirsutus et reste du genre (diphyes = dont la nature est double), non testée dans la présente étude
Figure 8. Aperçu de la composition du genre Dermanyssus à l’issue de la présente étude. Cette représentation fait écho à celle de la Fig. 1, p. 21. * espèces testées, appartenant à la faune commune en France. 6.3 Deux marqueurs complémentaires pour l’exploration intraspécifique Les gènes retenus pour l’exploration intraspécifique, la mt-Co1 et la Tropomyosine, sont aussi ceux qui aboutissent aux relations interspécifiques les mieux résolues (Publication V). Le degré de divergence entre populations au sein du genre Dermanyssus (D. gallinae au moins) est comparable entre ces deux gènes, pourtant situés dans des génomes dont la taille efficace de la population théorique est différente, et par conséquent dont la rapidité d'évolution diffère aussi. Le caractère fonctionnel de la portion majoritairement intronique de la Tropomyosine ayant des chances d’être très réduit comparé à celui des autres marqueurs testés, il n’est pas étonnant (1) qu’elle offre une image claire des relations phylogénétiques (faible taux de convergence probable), (2) qu’elle soit porteuse d’une information intraspécifique plus importante que les autres marqueurs nucléaires. En effet, s’il est vrai que la taille efficace de la population Ne est trois fois supérieure à celle des gènes mitochondriaux, l’effet de la sélection naturelle est quant à lui presque inexistant, la dérive génétique laisse probablement seule sa marque. Dans la mt-Co1, à l’inverse, celle-ci agit plus rapidement, de par la réduction de Ne, mais elle est contrainte par le rôle fonctionnel des différentes bases du codon. La théorie neutraliste de l’évolution moléculaire (Kimura 1968) explique partiellement l’équivalence constatée des divergences entre séquences de mt-Co1 et séquences de l’intron n de la Tropomyosine. Kimura défend en effet l’hypothèse qu’une grande majorité des polymorphismes génétiques moléculaires résulte de l’évolution par dérive génétique d’allèles mutants sélectivement neutres. La
241 vitesse d’évolution de la mt-Co1 aux mutiples positions contraintes par la fonctionnalité du codon, elle bénéficie d’une taille réduite de la Ne. A l’inverse, la réduction comparative de la vitesse d’évolution de l’intron n de la Tropomyosine est légèrement contrecarrée par sa nature non codante, donc plus à même d’être touchée par la dérive génétique. Cela n'est d'ailleurs pas le cas des ARN ribosomiques testés dans cette étude (séquence 18S-28S, incluant ITS1, 5,8S et ITS2), qui offrent en l'occurrence un très faible pourcentage de divergence dans le genre Dermanyssus. Cela n’explique toutefois pas une telle similitude des pourcentages de divergence entre un gène mitochondrial et un gène nucléaire au sein des populations de l'espèce D. gallinae. La mise en évidence d'une radiation suivie d'hybridations multiples par des analyses combinant phylogénie et génétique des populations ont permis d’expliquer cela (cf. plus bas). 6.4 Patterns écologiques révélés
a - Spécificité d’hôte La spécificité d’hôte dans le genre Dermanyssus, comme on peut s'y attendre dans un système macroproie / microprédateur plus encore que dans un système hôte / ectoparasite typique, apparaît régie en grande partie par des paramètres écologiques. Toutefois, des caractéristques intrinsèques du parasite jouent aussi un rôle important.
6.4.a.1 Paramètres écologiques Les habitudes de l’oiseau hôte influent fortement sur le spectre d’hôtes, puisque les transferts d’une espèce d’hôte à l’autre ont été révélés possibles en cas de partage du nid, c’est-à-dire réutilisation d’une saison sur l’autre. Ainsi D. hirundinis se montre-t-il en France cantonné aux Hirundinidae (Delichon urbica, Hirundo rustica), mais est aussi présent chez le moineau domestique (Passer domesticus) et le troglodyte mignon (Troglodytes troglodytes), les deux espèces d’oiseaux qui sont précisément connues pour réutiliser les nids de l’hirondelle rustique H. rustica. Le parallèle avec le spectre d’hôtes constaté aux Etats-Unis chez cette même espèce est emblématique : l’un des Hirundinidae parasités dans cette région du Monde (l’hirondelle des arbres, Tachycineta bicolor) étant cavernicole et habitué à faire usage de nichoirs fréquemment utilisés par le troglodyte familier T. aedon et la mésange à tête noire Parus atricapillus, D. hirundinis a pu être isolé chez ces trois espèces d’hôte. Or T. bicolor est absente du continent européen et aucune des mésanges testées en France ne réutilisent le nid des hirundinidae français. L’absence de D. hirundinis chez nos mésanges françaises du genre Parus, pourtant largement testées au cours de cette étude, confirme encore la nécessité de partage du nid pour le transfert d’un hôte à l’autre. Et cette possibilité est aussi soutenue par quelques essais de transfert artificiel au laboratoire de D. hirundinis sur canari. D. gallinae, D. carpathicus et D. longipes ont montré leur aptitude au transfert dans des conditions similaires. Cela s’accorde avec la capacité à survivre en l’absence d’hôte durant les mois d’absence in natura déjà démontrée chez D. gallinae et D. hirundinis (cf. §2.1.e.1). La capacité à transiter d’un nid à l’autre semble quant à elle fortement et rapidement limitée par la distance. Deux cas de nids d’hirondelles contenant des individus appartenant à D. hirundinis, construits à proximité de poules fortement parasitées par des populations de D. gallinae, soulignent une mobilité ou une capacité à détecter un hôte à distance réduites chez ces acariens. En effet, des cas d'espèces de ces microprédateurs cantonnées à leur hôte respectif, même en cas d’absence de leur hôte propre, ont mis en évidence cette faible mobilité ou aptitude à localiser un hôte.
242 6.4.a.2 Paramètres intrinsèques Un transfert sur un hôte nouveau semble possible chez ces espèces en cas d’opportunité dans un rayon d'action très réduit. Toutefois, une fois transféré artificiellement en conditions de laboratoire, D. carpathicus a montré à plusieurs reprises être incapable de se développer sur poule et l’installation de populations de la lignée PAS de D. longipes et de D. apodis sur canari a échoué (absence d’amplification de la population après plusieurs mois, mort de l’ensemble des individus). Certes, ces observations au laboratoire n’ont été faites pour la plupart qu’en dehors d’expérimentations strictement standardisées et planifiées. Elles manquent par conséquent de répétabilité et ne sont pas statistiquement exploitables. Toutefois, cela oriente la réflexion vers les caractéristiques intrinsèques qui président à la spécificité d’hôte dans le genre Dermanyssus. Des points communs et des différences entre les cinq espèces testées sont directement en relation avec les deux filtres génétiques définis par Combes (2000). Ces deux filtres représentent respectivement les caractéristiques impliquées (1) dans la rencontre entre parasite et hôte potentiel, (2) dans la compatibilité post-rencontre (nécessaire à la durabilité du système). En effet, parmi les cinq espèces explorées sur le plan écologique au cours de la présente étude, les points communs semblent relever du filtre 1, l’oiseau paraissant toujours être un vecteur à la condition sine qua non que le nid soit partagé (d’une saison sur l’autre, ou d’une couvée à l’autre), jamais de proche en proche. Des différences importantes en revanche relèvent du filtre 2. En effet, l’adaptabilité à des microécosystèmes plus diversifiés semble caractériser les lignées synanthropes et généralistes de D. gallinae, alors que les espèces basales ainsi que D. apodis (clade A mitochondrial), beaucoup plus spécialisées, sont confinées aux nids d’oiseaux sauvages, et apparemment aux oiseaux habitués à soigner au moins un peu l'hygiène du nid. Il n’est d’ailleurs pas inintéressant de noter que les différentes lignées de D. gallinae ont été recensées chez des oiseaux aux habitudes très diverses, mais ses populations étaient relativement faibles en nombre chez la plupart des oiseaux sauvages intégrés à l’analyse, si l’on compare aux explosions de population relevées couramment en conditions d’élevage. Les seuls cas de populations importantes dans l’avifaune sauvage, accompagnées d’ailleurs d’une prévalence très élevée, ont toutefois été remarqués chez l’étourneau d’Europe (Sturnus vulgaris), grâce à la colonie suivie et entretenue par le biais de plus de 30 nichoirs par l’ornithologiste J. Komdeur à Groningen (Pays-Bas). Or, parmi les espèces testées d’oiseaux ayant permis d’isoler des individus du genre Dermanyssus au cours de l’étude, l’étourneau d’Europe est de loin le plus ignorant des règles hygiéniques qui consistent à débarrasser le nid des fientes des poussins. La prolifération de D. gallinae en présence de fientes accumulées contraste avec le développement des autres espèces dans des nids « nettoyés ». D. apodis, cependant, espèce à l'écologie intermédiaire, parasite les martinets, dont les habitudes hygiéniques sont moindres que chez la plupart des autres espèces testées (mésanges, par exemple). En effet, les martinets communs adultes ingèrent les fientes des poussins durant une partie de la période de nidification, mais n'en évacuent pas (Dell’Omo et al. 1998). La nature de l’interaction entre fientes accumulées et sélection d’espèces du genre Dermanyssus est difficile à expliciter à ce stade de l’étude. Au laboratoire, D. carpathicus a montré des difficultés à se développer sur un canari à demeure, autorisé à déposer des fientes fraîches en permanence. L’entretien des souches de cette espèce nécessite le maintien du canari dans une cage séparée, son introduction dans le terrarium contenant les acariens n’ayant lieu qu’une nuit par semaine. Ainsi, les fientes s’accumulent en faible quantité et ont le temps de sécher. Les souches de D. gallinae, au contraire, se développent très bien dans une cage avec un canari à demeure. De nombreuses explications peuvent être envisagées, dont celles impliquant les possibles altérations de l’atmosphère induites par la présence de fientes en grande quantité : la composition de l’air, très
243 probablement affectée par ce paramètre, joue-t-elle un rôle direct sur le développement des parasites ? Des compétitions entre les espèces du genre Dermanyssus s’ajoutent-elles à une gêne plus ou moins importante engendrée par l’altération de l’atmosphère ambiante ? Celle-ci s’oppose-t-elle à l’installation de prédateurs ou de microorganismes entomopathogènes auxquels les lignées de D. gallinae seraient plus sensibles que les autres ? Ou encore le dégagement de certains gaz (ammoniac par exemple) pourrait-il entraver les éventuelles interactions par des phéromones entre acariens dans certaines espèces (confusion sexuelle, par exemple) ? Bien d’autres paramètres physico-chimiques ou mécaniques liés à l’accumulation de fientes non envisagés ici peuvent avoir un impact inattendu. Enfin, Clayton et Tompkins (1994) montrent que la virulence des ectoparasites est proportionnelle à la quantité de transmission horizontale. Suivant leur théorie, les ectoparasites capables de transmission indépendante sont extrêmement virulents alors que ceux qui dépendent du contact direct avec l’hôte ne sont pas virulents. Mais ils comparent des ectoparasites très différents entre eux tant phylogénétiquement qu’écologiquement (D. gallinae et poux mallophages). Si l’on considère les espèces du genre Dermanyssus étudiées ici, on s’aperçoit que la règle de Clayton et Tompkins (1994) trouve son application dans la mesure où l’on précise quelque peu la notion de transmission. En effet, si la virulence de D. gallinae et de D. prognephilus a pu être démontrée, celle de D. hirundinis n’est pas évidente (cf. 2.1b - ). Apparemment, le filtre 1, celui de la transmission, n’est pas différent entre les espèces testées du genre Dermanyssus. Mais le filtre 2 semble très différent. Il est probable que, si D. hirundinis peut transiter d’une hôte à l’autre en cas de partage de nid, sa compatibilité avec l’hôte soit beaucoup plus hasardeuse que celle des deux espèces ci-dessus. La théorie de Clayton et Tompkins semblerait donc bien s’appliquer au sein du genre Dermanyssus, entre espèces beaucoup plus comparables entre elles que leurs modèles. Mais il faudrait préciser que l’on doit considérer la quantité de transmission horizontale réussie (ou dont le succès est probable), c’est-à-dire sous l’angle à la fois du filtre 1 (transmission) et du filtre 2 (compatibilité post- transmission). L’apparente faiblesse de la virulence de D. hirundinis pourrait être corrélée à sa faible capacité à s’adapter en cas de transfert sur un nouvel hôte, même si cela ne lui est pas absolument impossible.
b - Transition sauvage-synanthrope : hybridation et radiation chez D. gallinae Sur la base de l’ensemble des analyses phylogénétiques, D. apodis se positionne comme une espèce intermédiaire entre les lignées synanthropes de D. gallinae (clade B mitochondrial ; cf. Publication IV) et les espèces typiquement rustiques que sont D. hirundinis, D. longipes, D. carpathicus, D. quintus et D. hirsutus (clade A mitochondrial, base de l’escalier de la topologie nucléaire). L’ancêtre commun étant par excellence adapté à l’avifaune sauvage, D. apodis, inféodée aux martinets, pourrait être le témoin d’une transition écologique vers la synanthropicité du clade D. gallinae, puisque ses hôtes de prédilection fondent très souvent leurs colonies en ville, dans des milieux fortement anthropisés, même s’ils demeurent dans des micro-écosystèmes d’oiseaux sauvages (nids). En outre, la lignée L1, dont les populations présentent un isolement reproducteur relativement avancé, parmi les autres lignées interhybridées de D. gallinae, parasite principalement les pigeons bisets, concurrents habituels des martinets quant aux sites de nidification (Nature Midi- Pyrénées, 2001). Cela semble confirmer ce statut transitoire dans l’évolution de l’écologie des espèces du groupe gallinae et, par là-même, le caractère intrinsèque de l’adaptabilité des lignées de D. gallinae aux milieux anthropisés.
244 Par ailleurs, plusieurs éléments sont en faveur d'une radiation ancienne au sein de Dermanyssus, qui pourrait être la conséquence d'activités humaines. Seehausen et al. (2004, 2007) ont montré qu'une radiation peut être une conséquence adaptative d’hybridations entre espèces naissantes, et que l’homogénéisation des milieux par l’action humaine (implantation de cultures, d’élevages, constructions diverses, …) peut en être la cause. Ces auteurs démontrent la défragmentation des milieux par le fait de l'homme peut engendrer une perte de la biodiversité : en effet, certaines barrières écologiques se voient brisées et des lignées récemment isolées les unes des autres sur le plan reproducteur (ie des espèces naissantes) peuvent se retrouver en contact. Ces entités spécifiques en formation n’ont souvent pas encore développé d’incompatibilité pré- ou post- zygotique, et peuvent s’hybrider avec succès. Cela non seulement interrompt les événements de spéciation en cours, mais encore, en produisant de nouvelles combinaisons de gènes, qui plus est dans un environnement brutalement altéré, la sélection naturelle intervient et une accélération du processus évolutif tend à en résulter. Cette accélération peut, selon Seehausen et al. (2004), donner lieu à des spéciations multiples et concomitantes (radiation). Or ce sont précisément les lignées de l’espèce synanthrope qui trahissent une radiation (Publication V). Et les événements de spéciations de cette radiation semblent avoir déjà largement avorté, puisque une forte réticulation entre isolats est manifeste et que de nombreux isolats – et individus – partagent des séquences de Tropomyosine très divergentes. Et les séquences mitochondriales, au taux de mutation par nature plus élevé sur une période donnée, bien que contraintes par le caractère fonctionnel des codons (mt-Co1) ou de la structure secondaire de l'ARN codé (16S), dénoncent une différenciation déjà bien avancée, tout au moins dans les isolats sauvages. Enfin, la méthode ABC (Publication V) met à jour davantage de flux de gene ancien que de flux de gène récent. Des hybridations multiples semblent donc avoir là encore suivi de près la radiation, sans doute rendues possibles par les mouvements humains, dont l’échelle et la fréquence dépasse de loin les mouvements des autres espèces (échanges internationaux, voire transcontinentaux). Malheureusement, l'importance des flux de gène rend la datation des divergences très difficiles, voire impossible, suivant la méthode ABC (cf. Publication V). La concordance entre action humaine et radiation, action humaine et hybridations demeure donc seulement supposée. Enfin, preuve de plus de cette succession ancienne radiation-hybridations, deux lignées non basales montrent encore un certain isolement reproducteur : isolement partiel de la lignée nommée L3 dans la publication V et isolement apparemment complet de la lignée spéciale L1 (publication III, V). La lignée L1 est étroitement liée au pigeon. La monophylie des isolats testés chez cette lignée est visible dans toutes les topologies (sauf avec la matrice des insertions/délétions de la Tropomyosine seule et sur la base des ITS). Sur la base des tests statistiques de différenciation (Fst) appliqués aux séquences de mt-Co1 et de Tropomyosine, l’isolat 9001 de la lignée spéciale L1 (Publication V) se montre moins différencié des autres isolats de D. gallinae que de ceux de D. apodis et D. hirundinis, mais davantage de chacun des autres isolats de D. gallinae que ces isolats ne le sont entre eux.
c - Structure de populations La structuration des populations au sein de D. gallinae apparaît en outre très différente entre faune sauvage et élevage. Les analyses utilisées ici ont porté uniquement sur un nombre réduit d'isolats largement explorés (plus de 20 individus séquencés par isolat). En effet, l’isolat sauvage de référence IL montre une équivalence entre valeurs attendues et valeurs observées de la diversité des séquences (nombre d’haplotypes et diversité haplotypique en fonction du nombre de séquences testées dans l’isolat et du nombre de sites ségrégeants). Les isolats de D. gallinae prélevés dans des élevages, en revanche, révèlent des déséquilibres importants entre diversité attendue et diversité
245 observée dans les deux gènes (nombre d’haplotype et diversité haplotypique très inférieurs aux valeurs estimées). Cela peut témoigner soit événements fondateurs, soit d’une simple mise en contact récente de populations isolées les unes des autres. Le premier cas est commun chez les espèces déprédatrices en général, aussi bien phytophages qu’hématophages ou envahissant les denrées stockées, dont l’homme cherche à contrôler les populations par divers moyens (pesticides naturels ou de synthèse, organismes auxiliaires, …). Une réduction brutale de la taille de la population par l’application de ces moyens réduit fortement la diversité génétique (goulot d’étranglement) et, sélectionnant aléatoirement les génotypes, induit un déséquilibre des rapports de diversité haplotypique. Après une phase de développement à partir de ce nombre réduit d’individus, la population peut retrouver sa taille originale, mais est marquée génétiquement. Quant au second cas possible, le mélange récent de populations isolées, l’action de l’homme peut là aussi en être la cause directe : les échanges commerciaux au sein des filières d’élevages sont tout à fait susceptibles d’initier accidentellement un contact entre des populations d’acariens établies dans des élevages distants, depuis longtemps, par le transfert des animaux et du matériel d’élevage. Le déséquilibre très significatif dans les élevages ne s’explique sans doute pas de la même manière pour les deux gènes, dont les vitesses d’évolution sont très différentes par nature, et par conséquent, témoins d’événements très décalés dans le temps. Le déséquilibre des valeurs pour les séquences de Tropomyosine, qui offrent une image relativement ancienne, important au sein des isolats de tous les élevages de poules testés, s’accorde aisément avec un goulot d’étranglement concomitant avec la domestication des volailles ou la mise en place de pratiques de lutte dans les élevages. En effet, tous les isolats provenant d’élevage de poules présentent cette caractéristique. L’isolat 9001, en provenance d’un élevage de pigeon, représentant de la lignée spéciale L1, possède un seul haplotype de Tropomyosine, ce qui exclut toute interprétation quant au caractère prévisible ou non de cette faible diversité. Il n'en demeure pas moins qu'il est susceptible de représenter l'état "naturel" des lignées originelles de D. gallinae, c'est-à-dire une séquence de Tropomyosine très stable, comme chez les autres espèces testées dans le genre Dermanyssus. En revanche, le déséquilibre dans la diversité des haplotypes de mt-Co1 n’est pas relevé dans tous les élevages de poules avec le même degré de significativité : l’isolat SK, prélevé au Danemark il y a 12 ans et confiné en laboratoire depuis, ne présente pas de déséquilibre significatif entre valeurs observées et valeurs attendues. Il porte en outre la marque d’un isolement évident par rapport aux isolats provenant directement d’élevages français. Une alternative se présente comme suit : soit les isolats français ont vécu un second événement fondateur, très postérieur au premier, soit le déséquilibre des isolats français rend compte d’un simple mélange de population par les échanges commerciaux. Le confinement de SK l’aurait soustrait à l’effet du contrôle des parasites ou de l’introduction de populations au matériel génétique très différent. Etant donnée l’ampleur de la divergence entre séquences de mt-Co1 isolées dans les élevages de pondeuses, si l’on compare à SK, le mélange de populations isolées semble nettement la plus plausible des interprétations. Le rôle des flux commerciaux semble donc important dans la dissémination des acariens au sein de la filière pondeuse francaise. 6.5 Du caractère invasif de D. gallinae et d’une espèce peut-être concurrente D. gallinae présente de nombreuses particularités qui l’opposent aux autres (synanthropicité, variabilité génétique et morphologique accrue, faiblesse de la spécificité d’hôte …) et en font un candidat redoutable pour l’invasion des milieux avicoles (cf. Publication V). D’une manière générale,
246 les analyses utilisant chacune des deux portions de gène utilisées révèlent de fortes différences entre D. gallinae et les autres espèces testées, tout au moins D. apodis et D. hirundinis. La diversité des haplotypes, tant en nombre qu’en degré de divergence y est très nettement plus élevé que chez les autres. Cela peut être le signe d’une architecture génétique différente entre l’entité D. gallinae et les autres membres du genre Dermanyssus testés, et par conséquent de potentielle variance génétique additive et/ou épistasie augmentées (Lee 2002). Les hybridations décrites plus haut ont pu contribuer à construire ces caractéristiques. L’espèce D. gallinae se présente comme un complexe de lignées en effervescence, s’isolant les unes des autres, amorçant des spéciations qui sont ensuite interrompues par des hybridations incessantes, le tout marquant une évolution accélérée et une flexibilité bien supérieure à ses sœurs spécialistes. Ces particularités intrinsèques associées à de nombreuses preuves d’une adaptabilité remarquable et couronnées de succès (spectre d’hôtes large, synanthropicité, expansion en cours constatée dans certaines zones du monde) clament le caractère fondamentalement apte à l’invasion de cette espèce. Le tableau réunit la plupart des caractéristiques recensées par Lee (2002), propres à ces rares espèces capables de réussir toutes les étapes de l’invasion décrites par Williamson (1996) : importation, introduction, établissement, développement nuisible. Et ce n’est pas un hasard si ce micropredateur généraliste se présente comme une entité dérivée dans un groupe de microprédateurs dont l’état plésiomorphe de la spécificité d’hôte est l’état spécialiste. Une fois encore, le caractère spécialiste pour un parasite n’est pas un cul de sac évolutif (Desdevises et al. 2002) et l’augmentation de la spécificité d’hôte n’est pas nécessairement un progrès. Curieusement, d’ailleurs, un motif comparable semble émerger d’une reconstruction phylogénétique proposée par Reinhardt et Siva-Jothy (2007) sur la base de caractères morphologiques et de traits d’histoire de vie, pour les Cimicidae : les punaises de cette famille, microprédatrices aptères comme Dermanyssus, montrent un spectre d’hôtes variable, quoique en général relativement restreint. Comme chez Dermanyssus, les entités basales dans la reconstruction de ces auteurs, tels les Primiciminae et les Latrociminae, montrent un spectre d’hôtes étroit. Et comme chez Dermanyssus, quelques espèces dont les synanthropes (genre Cimex, incluant les punaises de lit), en position distale, semblent être aptes à transiter d’un hôte à l’autre relativement aisément (chiroptères, homme, oiseaux). D. gallinae n’est toutefois pas adaptable à tous les environnements, puisqu’on le rencontre rarement dans les nids d’oiseaux aux habitudes hygiéniques développées (éjection des sacs fécaux des poussins hors du nid par les mésanges, par exemple ; cf. plus haut). Cependant, cette différence de tolérance vis-à-vis de l’accumulation des fientes fraîches n’est probablement pas la cause unique de l’absence de D. gallinae des nids de mésanges in natura puisqu’il est présent dans les nids de mésanges en verger biologique et intégré (Publication VI). Il est aussi probable que d’autres arthropodes, dont de potentiels prédateurs de D. gallinae soient gênés par l’accumulation de fientes (et/ou par certains traitements utilisés en verger biologique et intégré). En outre, D. gallinae paraît absent des nids d’oiseaux en agroécosystème de verger conventionnel (i.e. traités chimiquement). En revanche, une espèce potentiellement concurrente serait à explorer plus avant : O. sylviarum. Très présente dans l’avifaune sauvage en France - davantage même que D. gallinae selon nos résultats - cette espèce semble en outre relativement peu sensible à certains traitement phytosanitaires réalisés dans les vergers conventionnels (cf. Publication VI, pp. 214 sqq.). Or cette espèce a déjà montré son aptitude à s’adapter à certaines molécules de synthèses telles des organophosphorés et carbamates (Mullens et al. 2004). Par ailleurs, certains indicateurs d'une émergence en cours de cette espèce de Macronyssidae en conditions d’élevages commencent à apparaître en France : isolement d’individus dans un élevage amateur de canaris (Rhône) et dans un élevage de faisans (Vendée), témoignages d’autres éleveurs de canaris et de faisans du Rhône évoquant fortement l’infestation à Macronyssidae (« pou rouges » visibles sur l’hôte), et ce, apparemment depuis deux ou trois années. Les échantillonnages réalisés au
247 cours de la présente thèse dans plusieurs dizaines d’élevages de pondeuses de France de types variés (au sol/en cages, conventionnels/biologiques) ont permis de détecter uniquement des acariens appartenant à D. gallinae. L’annonce de la présence d’O. sylviarum dans des élevages de pondeuses français des mêmes régions par Bruneau et al. (2002) semble curieuse. Quoi qu'il en soit, les résultats de la présente thèse soulignent qu’aujourd’hui seul D. gallinae pose des problèmes en élevage de pondeuses français. La présente étude a mis en évidence la flexibilité remarquable de D. gallinae, mais n’a pas travaillé sur celle d’O. sylviarum. Or à l’heure actuelle, les deux espèces distantes entre elles phylogénétiquement présentent des répartitions écologiques symétrique si l'on considère les Etats- Unis et l'Europe. Occupant des niches chevauchantes, si ce n’est similaires, elles sont présentes dans les faunes sauvages nord américaine et française et absente la première des élevages de pondeuses nord américaines, la seconde des élevages de pondeuses français. Cela est-il dû à de grandes différences dans les moyens de production (et par conséquent dans l’écologie des microécosystèmes concernés) ? Ou des invasions de l’un et/ou de l’autres dans les élevages des régions opposées sont- elles en cours chez ces espèces ? Quelques récents témoignages d’élevages de volailles d’Europe du Nord contaminés par O. sylviarum (Chirico, communication personnelle) laissent penser que cette espèce est en cours d’expansion dans ces milieux dans l’Ancien Monde. Les changements climatiques y sont-ils pour quelque chose ? 7 Conclusions et perspectives
7.1 Conclusions sommaires La première partie de thèse retrace un cheminement d’ordre taxinomique et a permis de poser les bases nécessaires à la seconde partie, ainsi qu’à toute étude d’ordre écologique d’espèces du genre Dermanyssus communes dans l’avifaune francaise au moins. De ce long préalable ont émergé plusieurs informations nouvelles, et certaines des connaissances antérieures ont pu être confirmées et précisées. La base taxinomique qui en résulte a permis (1) d’obtenir un inventaire des espèces de Dermanyssus présentes en France, (2) de commencer à explorer certaines caractéristiques écologiques du groupe gallinae en relation avec sa phylogénie (spécificité d’hôte, flexibilité évolutive).
a - Nouveautés taxinomiques Un aperçu synoptique des remaniements taxinomiques est présenté dans la Figure 8 (p. 241), à comparer avec l’aperçu initial de la Figure 1 (p. 21). Contrairement à notre première hypothèse, aucune espèce n’a pu être mise en synonymie. Seule D. gallinoides semble représenter un synonyme junior de D. gallinae, mais l’échantillonnage sur Picidae (sa famille d’hôtes type) est insuffisant pour l’affirmer avec certitude. Les autres espèces sont nettement isolées les unes des autres. En outre, une espèce a été décrite consécutivement au travail de délimitation des espèces (D. apodis), une lignée de D. gallinae constitue aussi sans doute une espèce inédite, quoique le caractère définitif de son isolement reproducteur reste à vérifier, et D. longipes regroupe manifestement deux entités spécifiques différentes. La diversité du genre Dermanyssus est apparemment plutôt sous-estimée à l’heure actuelle que surestimée, comme nous le pensions au début de l’étude.
b - Nouveautés écologiques Sur la base taxinomique obtenue, le simple inventaire des espèces du genre Dermanyssus dans des échantillons provenant de l’avifaune sauvage (principalement française) a révélé une étroitesse
248 inattendue du spectre d’hôtes chez certaines espèces. Parmi les deux espèces réputées les moins spécifiques du groupe, D. hirundinis s’avère parasiter un spectre d’hôtes nettement réduit, au moins en France (une seule famille, celle des Hirundinidae), alors que D. gallinae confirme sa faible spécificité. Quant à notre nouvelle espèce D. apodis, elle semble bien être inféodée aux martinets (genre Apus). La clarification taxinomique était donc bien nécessaire à un réajustement de l’appréhension de la spécificité d’hôte dans ce genre. Comme McCoy et al. (2003) l’on montré au sein de l’espèce I. uriae, la biologie de l’oiseau hôte semble jouer un rôle important dans la dispersion des populations d’acariens. Tout nidicole qu’il soit, et si rapide que soit son repas (cf. § Avant- propos), le microprédateur du genre Dermanyssus semble bien profiter des déplacements de son hôte ailé pour se déplacer, de la même manière que la tique I. uriae. Les espèces microprédatrices du genre Dermanyssus ne demeurent pas longtemps sur l’hôte pour la prise du repas, mais il semblerait qu’un essaimage « volontaire » puisse avoir lieu par l’intermédiaire des femelles adultes : il est noté dans la publication III que les quelques individus prélevés directement sur un hôte hors du nid au cours de la présente étude étaient presque systématiquement des femelles. Un constat similaire a été établi par plusieurs auteurs chez des Cimicidae (Reinhardt et Siva-Jothy 2007). Par ailleurs, le rôle des flux commerciaux dans la dispersion des populations de D. gallinae en élevage de pondeuses s’avère primordial, suggérant une capacité à voyager par l’intermédiaire d’objets divers, comme cela a déjà été noté par Reinhardt et Siva-Jothy chez les punaises de lit (2007). Enfin, un autre parallèle peut être établi avec les Cimicidae (incluant les punaises de lit), dont les habitudes plésiomorphes semblent plutôt être une spécificité d’hôte relativement élevée (Reinhardt et Siva-Jothy 2007).
Des différentes analyses phylogénétiques menées au cours de la présente thèse, il ressort en outre une forte opposition entre D. gallinae et les 4 autres espèces françaises testées : (1) D. gallinae manifeste une synanthropicité marquée, à l’opposé des autres espèces françaises, beaucoup plus rustiques, (2) D. gallinae est largement généraliste, tandis que les autres espèces françaises sont spécifiques (à des niveaux variés), (3) D. gallinae présente une flexibilité et un potentiel évolutifs très accrus, comparé aux autres espèces. Or ces différences semblent résulter non seulement de l’activité humaine et de ses retombées environnementales, mais aussi, pour une grande part, de caractéristiques intrinsèques. D. gallinae, complexe d’espèces en ébullition et en position distale dans les topologies nucléaires, apparaît fondamentalement équipé pour une invasion des milieux avicoles. Composée de lignées qui ont rapidement divergé les unes des autres et se sont hybridées entre elles, l’architecture génétique de cette espèce la rend sans doute fortement apte à s’adapter rapidement à des modifications dans les conditions écologiques dans lesquelles elle vit. La phylogénie proposée par Reinhardt et Siva-Jothy (2007) présente les espèces de Cimicidae infestant l’homme (Cimex lectularius et C. hemipterus) comme en position distale (avec d’autres). Un motif comparable est-il envisageable chez cet hémiptère aux mœurs très proches (microprédateur aptère, cf. § Avant-propos), et en pleine réémergence dans divers pays, par exemple aux Etats-Unis (Szalanski et al. 2008), en Corée (Lee et al. 2008) ? Toutefois, un comportement fréquent d’accouplement interspécifique entre les deux espèces citées ci-dessus, apparemment très proches, a prouvé être délétère par l’absence d’hybridation qui en est issue (œufs stériles) (Newberry 2008). Les fréquentes hybridations démontrées chez D. gallinae ne sont peut-être pas possibles chez les espèces du genre Cimex. Par ailleurs, une espèce de Macronyssidae, O. sylviarum, l’équivalent nord américain de D. gallinae en élevage de pondeuses, représente peut-être une concurrence non négligeable. D’autant que les analyses de nids réalisés au cours de la présente thèse ont permis de révéler une situation symétrique en Europe et au Etats-Unis : D. gallinae est présent en condition sauvage comme en élevage de pondeuse en France, seulement (ou presque) en condition sauvage aux Etats-Unis, et vice
249 versa pour O. sylviarum. Mais cette dernière espèce a déjà été signalée en élevage de pondeuses en Europe du Nord, et est actuellement rencontrée dans des élevages très différents en France (faisans, canaris). 7.2 Perspectives Comme souvent à l’issue d’un travail de recherche, si quelques éléments de réponse à certaines questions ont pu être formulées, bien davantage de questions nouvelles ont été soulevées. Nombreuses sont les perspectives de recherche qui peuvent découler de la présente étude, et nombreuses aussi sans doute les questions à « trouver ». Voici les principales :
a - Exploration des flux de populations au sein de l’espèce D. gallinae Une analyse utilisant les outils de la génétique des populations intégrant un plus grand nombre d’isolats massivement séquencés que la publication V a été entamée, et vise à clarifier les voies de dissémination entre populations infestantes en élevages de pondeuses en Europe, voire à l’échelle mondiale. Une publication devrait en émerger, dans le prolongement de l’exploration intraspécifique esquissée dans la publication V. Ce travail en cours de réalisation est financé par le Pôle d’Expérimentation et de Progrès (PEP) avicole de la Région Rhône-Alpes, et mené en collaboration avec l’Institut Technique de l’AVIculture (ITAVI) (Sophie Lubac).
b - Espèces cryptiques Le statut spécifique de la lignée L1 (D. gallinae) et des deux lignées regroupées dans D. longipes est à confirmer, et le cas échéant, deux espèces cryptiques sont à décrire. Pour cela, des analyses, morphologiques et moléculaires sont nécessaires, de même, si possible que des essais de croisement au laboratoire.
c - Analyse moléculaire et morphologique de l’ensemble du genre Dermanyssus Une analyse complète (morphologie et marqueurs moléculaires) impliquant un plus grand nombre des espèces actuellement décrites permettrait de peaufiner certaines délimitations interspécifiques non explorées ici au sein du genre Dermanyssus, en particulier chez les espèces que Moss classait dans le groupe hirsutus du sous-genre Dermanyssus et dans le sous-genre Microdermanyssus. Selon notre analyse impliquant 20 des espèces actuellement décrites, ces deux groupes n’en formeraient qu’un. Toutefois, l’inclusion de caractères moléculaires ayant montré une efficacité plus importante pour la résolution des relations intragénériques parmi les espèces testées, il serait intéressant de voir s’ils confirment la monophylie du groupe hirsutus de Moss, ainsi que leur insertion parmi les espèces du groupe gallinae sensu Moss (1978). En outre, l’intégration d’une ou deux espèces du genre Liponyssoides permettrait de tester réellement le monophylie du genre Dermanyssus.
Par ailleurs, une investigation plus poussée de la région de l’EF1-D séquencée au cours de l’étape liminaire (cf. § Remarques sur la publication III pp.) pourrait permettre de comprendre le problème rencontré. S’agit-il de paralogie, avec une double copie dont une ou l’autre éteinte chez certaines espèces ? Un séquençage massif d’un plus grand nombre de populations et leur analyse phylogénétique suivie d’une réconciliation des arbres de gène avec les arbres d’espèces tels que nous les avons fixés dans la présente étude pourrait faire émerger une explication à l’incongruence
250 rencontrée (Page & Charleston 1997) et apporter un complément d’information quant aux relations phylogénétiques dans le groupe gallinae.
d - Cophylogenèse au niveau population Enfin, la spécificité d’hôte s’est avérée plus importante que l’on ne le pensait auparavant. Des facteurs écologiques semblent jouer un rôle important dans le transfert d’un hôte à l’autre, mais des caractéristiques intrinsèques sont aussi très impliquées. La forte opposition entre les habitats permettant la prolifération des lignées de D. gallinae et ceux hébergeant les autres espèces françaises ainsi que le rôle vecteur de l’oiseau mis en évidence dans la publication IV témoignent des premiers. Les caractéristiques évolutives particulières mises à jour chez D. gallinae par comparaison avec D. hirundinis et D. apodis dans la publication V signent les secondes. Ces éléments ne sont toutefois que les témoins de particularités dans la relation hôte-parasite chez ces espèces du groupe gallinae qui demeurent au moins partiellement inconnues. Une approche cophylogénétique au niveau population tant pour les acariens que pour leurs oiseaux hôtes, associée à des bioessais sur le terrain, permettrait une appréhension approfondie de ces aspects.
Dans cet esprit, deux projets ont été soumis à l’Agence Nationale de la Recherche (ANR), mais n’ont pas été retenus. Leur mise en œuvre pourrait aboutir à une meilleure compréhension de ce qui préside à la dissémination et au développement des populations des espèces du groupe gallinae.: Avipred - « Spécificité d’hôte chez un ectoparasite nidicole : association hôte-parasite chez quelques espèces microprédatrices de Dermanyssus » – ANR Blanc édition 2009 Phylopred – « Spécificité d’hôte chez un ectoparasite nidicole : cophylogénie hôte-parasite chez quelques espèces microprédatrices de Dermanyssus » – ANR Jeunes Chercheuses et Jeunes Chercheurs édition 2009
e - Investigation de la situation symétrique entre D. gallinae et O. sylviarum en France et aux Etats-Unis Une exploration utilisant les outils de la phylogénie et de la génétique des populations visant à comparer une nombre important d’isolats des deux espèces provenant des deux pays, et des deux grands types d’environnements (sauvage, élevage) pourrait permettre, en collaboration avec la filière avicole, de détecter des raisons écologiques liées aux pratiques d’élevages des deux pays dans le déséquilibre constaté entre les deux espèces et/ou de mettre en évidence une expansion actuelle de l’une ou des deux espèces dans les zones respectivement non colonisées. Cela pourrait aboutir à terme à des préconisations à l’attention des éleveurs et/ou des fabricants de structure d’élevage en vue d’une prophylaxie améliorée. Cela est en outre susceptible d'offrir un aperçu de certains effets non envisagés actuellement dans les élevages de volaille des changements globaux.
f - Comparaison des valeurs de polymorphisme de séquences d’ADN nucléaires et mitochondriales entre espèces de microprédateurs aptères Il serait intéressant de voir si le motif de la diversité haplotypique augmentée, et éventuellement des hybridations multiples, chez l’espèce synanthrope et généraliste D. gallinae, contrastant avec les données des autres espèces du genre se répète chez des espèces aux mœurs similaires : explorer ainsi O. sylviarum vs autres Macronyssidae, C. lectularius et C. hemipterus vs autres Cimicidae pourrait permettre d’appréhender ce qui préside au développement de ces espèces microprédatrices synanthropes, et peut-être d’assouplir et affiner des règles trop rigides que l’on risquerait de tirer du cas Dermanyssus.
251 8 Lexique arrhénotoque (parthénogenèse) (grec arrhên, mâle, et tokos, action d’enfanter, de mettre bas) : cas de la reproduction haplodiploïde fréquente chez certains arthropodes, dont les hyménoptères sociaux, où les œufs non fécondés donnent naissance aux mâles (haploïdes, issus de parthénogenèse) et les œufs fécondés donnent naissance aux femelles (diploïdes, issues de reproduction sexuée). bande (contexte : élevage) : ensemble des volailles maintenues en commun dans un type d’élevage entre leur entrée et leur réforme (retrait de l’élevage et abattage). Pour les pondeuses, l’entrée de la bande se fait aux alentours de 18 semaines (âge de début de mâturité pour la ponte) et la réforme, en conditions normales, correspond à la fin de la période optimale/rentable de ponte et a lieu environ un an plus tard. gonotrophique (grec gonos, semence génitale, et trophê, action de nourrir): un cycle gonotrophique désigne la succession recherche de l’hôte-repas de sang-maturation des œufs- oviposition caractérisant chaque ponte chez les femelles d’arthropodes hématophages. haplodiploïdie : cas des différents types de reproduction où les mâles sont haploïdes et les femelles diploïdes (selon Cruickshank et Thomas 1999). opisthosome (grec opisthe, ensuite, derrière, soma, corps) : partie du corps des acariens située à l’arrière des dernières paires de pattes. Equivalent de l’abdomen des insectes et Aranea (araignées, scorpions, …), l’opisthosome ne présente pas la séparation nette par rapport à la partie antérieure que l’on observe entre l’abdomen et le thorax chez les premiers, entre l’abdomen et le céphalothorax chez les seconds. organophosphoré : groupe de molécules à action neurotoxique visant l’acétylcholinestérase, médiateur chimique impliqué dans la transmission de l’influx nerveux. De nombreux organophosphorés sont utilisés comme insecticides ou acaricides en agriculture aussi bien végétale qu’animale. pseudoarrhénotoquie (cf. arrhénotoque) : cas particulier de l’haplodiploïdie* où les mâles sont issus d’œufs fécondés, mais à un stade ou un autre de l’embryogenèse desquels, l’élimination d’une moitié du génome, aboutit a posteriori à un état haploïde. vide sanitaire (contexte : élevage) : d’une durée d’env. 2 mois (souvent réduit à 3 semaines), le vide sanitaire est réalisé entre 2 bandes. En l’absence des volailles, un certain nombre de mesures à visée sanitaires sont appliquées (nettoyage, désinfection, désinsectisation). thigmotactisme (grec thigein, toucher et taxis, arrangement): immobilisation provoquée par le contact d’un corps solide, tendance à rechercher le contact de surfaces dures avec son corps. Le thigmotactisme amène certains animaux à se regrouper dans des espaces réduits. 9 Références bibliographiques Amin OM, Sonenshine DE. 1969. Development of the American Dog Tick, Dermacentor variabilis, Following Partial Feeding by Immatures. Annals of the Entomological Society of America, 63(1):128-133. Anderson RR and Harrington LC. 07/04/2009. MOSQUITO BIOLOGY for the HOMEOWNER. http://www.entomology.cornell.edu/MedEnt/MosquitoFS/MosquitoFS.html Axtell RC, Arends JJ. 1990. Ecology and Management of arthropod pests of poultry. Annual review of entomology 35:101-126. Banks JC, Paterson AM. 2005. Multi-host parasite species in cophylogenetic studies. International Journal for Parasitology. 35(7):741-746
252 Burtt EH Jr, W Chow, and GA Babbitt. 1991. Occurrence and demography of mites of Tree Swallow, House Wren, and Eastern Bluebird nests. Pages 104-122 in J. E. Loye and M. Zuk, editors. Bird- parasite interactions: ecology, evolution, and behaviour. Oxford University Press, Oxford, UK. Clayton, D. H., and D. M. Tompkins. 1994. Ectoparasite virulence is linked to mode of transmission. Proceedings of the Royal Society of London 256:211-217. Clayton, D. H., and D. M. Tompkins. 1995. Comparative effects of mites and lice on the reproductive success of Rock Doves (Columba livia). Parasitology 110:195-206. Cruickshank, R.H., 2002. Molecular markers for the phylogenetics of mites and ticks. Systematic and Applied Acarology 7:3-14. Cruickshank RH et Thomas RH. 1999. Evolution of Haplodiploïdy in Dermanyssine mites (Acari: Mesostigmata). Evolution 53(6):1796-1803. Dell’Omo G, Alleva E, Carere C. 1998. Parental recycling of nestling faeces in the common swift. Animal Behaviour 56(3):631-637 Desdevises Y, Morand S and Legendre P. 2002. Evolution and determinants of host specificity in the genus Lamellodiscus (Monogenea). Zoological Journal of the Linnean Society 77(4):431-443. Djernaes M. and Damgaard J. 2006. Exon-intron structure, Paralogy and Sequenced Regions of Elongation Factor-1 alpha in Hexapoda. Arthropod Systematic and Phylogeny 64(1):45-52. Dotson E. 1982. The question of para-haploidy or haplodiploidy in the chicken mite, Dermanyssus gallinae. M.S., Georgia Southern College, Statesboro. M.S. Thesis. Dowling APG. 2006a. Systematics, genomics, and the evolution of parasitism within Dermanyssoidea (Mesostigmata). XIIth International Congress of Acarology, Amsterdam, 21-26 août 2006. Dowling APG. 2006b. Mesostigmatid mites as parasites of small mammals: Systematics, ecology, and the evolution of parasitic associations. In Morand, Krasnov, and Poulin (eds). Micromammals and macroparasites : From evolutionary ecology to management. Springer-Verlag, Tokyo. Dunn LH. 1924. Life history of the Tropical Bed Bug, Cimex rotundatus, in Panama. The American journal of tropical medicine and hygiene S1-4(1):77-83. el Kady GA, Shoukry A, Ragheb DA, el Said AM, Habib KS, Morsy TA. 1995. Mites (acari) infesting commensal rats in Suez Canal zone, Egypt. Journal of the Egyptian Society of Parasitology 25(2):417-25. Evans GO., 1963. Some observations on the chaetotaxy of the pedipalps in the Mesostigmata (Acari). Annals & magazine of natural history 13, 513-527. Evans GO and WM Till. 1966. Studies on the British Dermanyssidae (Acari: Mesostigmata). Part II. Classification. Bulletin of the British Museum (Natural History) Zoology 14:109-370 Ewing, HE. 1923.The dermanyssid mites of North America. Proceedings of the United States National Museum 62: 1-26 pls 1, 2. Fain A, Galloway TD. 1993. Mites (Acari) from nests of sea birds in New Zealand. Ii. Mesostigmata and Astigmata. Bulletin De L'Institut Royal Des Sciences Naturelles De Belgique, Entomologie 63, p. 95-111 Fenća P, E Schniererová. 2004. Mites (Acarina: Mesostigmata) in the nests of Acrocephalus spp. and in neighbouring reeds. Biologia (Bratislava) 59: 41-47. Fenća P, E Schniererová. 2005. Mites (Acarina, Gamasida) in littoral zone of Jakubov fishponds (Slovakia), p. 9-14. In: Tajovský, K., Schlaghamerský, J., Pižl, V. (eds), Contributions to Soil Zoology in Central Europe I, Institute of Soil Biology Academy of Sciences of the Czech Republic, ýeské BudČjovice. Gorrochotegui-Escalante N, De Lourdes Munoz M, Fernandez-Salas I, Beaty BJ and Black IV WC. 2000. Genetic isolation by distance among Aedes aegypti populations along the northeastern coast of Mexico. The American journal of tropical medicine and hygiene 62(2):200-209.
253 Gu YM, Ting QY. 1992. New species and new records of Dermanyssus from China (Acari: Dermanyssidae). Acta Zootaxonomica Sinica 17:1:32-36. Guy JH ; Khajavi M ; Hlalel MM ; Sparagano 2004. Red mite (Dermanyssus gallinae) prevalence in laying units in Northern England. British Poultry Science 45:15-16 Hallan J. 2005. http://insects.tamu.edu/research/collection/hallan/Acari/Family/Mesostigmata1.htm. 04/07/2009. Hirst S. 1913. On three new species of Gamasid mites found on rats. Bulletin of entomological research 4: 119-124. Hoffmann G. 1987. Vogelmilbe als Lästlinge, Krankheitserzeuger und Vektoren bei Mensch und Nutztier. Deutsche Tierarztliche Wochenschrift 95:7-10. Hutcheson HJ et Oliver JH Jr. 1988. Spermiogenesis and Reproductive Biology of Dermanyssus gallinae (De Geer) (Parasitiformes: Dermanyssidae). Journal of Medical Entomology 25:5:321-330. Izri A. 2001. Les poux : diagnostic, nuisance et rôle vectoriel. Revue Française des Laboratoires 338:37- 40. Ives AR, Klug JL and Gross K. 2000. Stability and species richness in complex communities. Ecology letters 3:399-411. Johnson, L. S., and D. J. Albrecht. 1993. Effects of haematophagous ectoparasites on nestling House Wrens, Troglodytes aedon: who pays the cost of parasitism? Oikos 66:255-262. Keçeci T, Handemir E, Orhan G (2004) The Effect of Dermanyssus gallinae Infestation on Hematological Values and Body Weights of Cocks. Trkiye Parazitoloji Dergisi 28:192-196. Kilpinen O. 1999. Problems caused by the chicken mite, Dermanyssus gallinae, in the Danish egg production. Cost 833 Agriculture, 2. Annual Workshop, Cluj, Roumanie, pp. 61-65. Kilpinen O. 2001. Activation of the poultry red mite Dermanyssus gallinae (Acari : Dermanyssidae) by increasing temperatures. Experimental and Applied Acarology 25:859-867. Kilpinen O. 2005. How to obtain a bloodmeal without being eaten by a host: the case of poultry red mite Dermanyssus gallinae Physiological Entomology 30:232-240 Kilpinen O, Roepstorff A, Permin A, Nørgaard-Nielsen G, Lawson LG & Simonsen HB. 2005. Influence of Dermanyssus gallinae and Ascaridia galli infections on behaviour and health of laying hens (Gallus gallus domesticus). British Poultry Science 46 (1), 26-34. Kirkwood AC. 1967. Anaemia in poultry infested with the red mite Dermanyssus gallinae. Veterinary Record 29;80(17):514-6. Kirkwood AC. 1963. Longevity of the Mite Dermanyssus gallinae and Liponyssus sylviarum. Experimental Parasitology 14:358-366. Koehler PG, Pereira RM, Pfiester M, and Hertz J. 2008. Bed Bugs and Blood-Sucking Conenose. ENY- 227 (IG083), one of a series of the Entomology and Nematology Department, Florida Cooperative Extension Service, Institute of Food and Agricultural Sciences, University of Florida. Kuris AM and Lafferty KD 2000. Parasite-host modelling meets reality: adaptive peaks and their ecological attributes. In: Evolutionary Biology of Host-Parasite Relationships: Theory Meets Reality, Ed. Poulin R., Morand S. and Skorping A. Elsevier Science B.V. pp. 9-26. Lainson R. 1960. The Transmission of Lankesterella (= Atoxoplasma) in Birds by the Mite Dermanyssus gallinae. Journal of Eukaryotic Microbiology 7(4):321-322. Lee CE. 2002. Evolutionary genetics of invasive species. Trends in Ecology and Evolution 17(8):386-391 Lee IY, Ree HI, An SJ, Linton JA, Yong TS. 2008. Reemergence of the bedbug Cimex lectularius in Seoul, Korea. Korean Journal of Parasitology 46:269-71. Lopes J, Beaumont M. 2008. Approximate Bayesian Computation - Studying demographic parameters. Mathematics and Informatics in Evolution and Phylogeny, Montpellier 10-12 juin.
254 Majka CG, Klimaszewski J and Lauff RF. 2006. New Coleoptera records from owl nests in Nova Scotia, Canada. Zootaxa 1194:33-47. McCoy. 2001. Conséquence de la dispersion dans les systèmes hôtes-parasites : dynamique des populations, structure génétique, et adaptation locale chez un parasite d'oiseaux marins, la tique Ixodes uriae. Thèse nouveau doctorat. McCoy KD, Boulinier T, Tirard C, Michalakis Y. 2003. Host-dependent genetic structure of parasite populations: differential dispersal of seabird tick host races. Evolution 57(2):288-296. McDevitt R, Nisbet AJ, Huntley JF. 2006. Ability of a proteinase inhibitor mixture to kill poultry red mite, Dermanyssus gallinae in an in vitro feeding system. Veterinary Parasitology 5:141(3-4):380- 5. Merkl O., Bagyura J. and Rózsa L. 2004. Insects inhabiting Saker (Falco cherrug) nests in Hungary. Ornis Hungarica 14:1-4. Meyer-Kühling B, Pfister K, Müller-Lindloff J, Heine J. 2007. Field efficacy of phoxim 50% (ByeMite®) against the poultry red mite Dermanyssus gallinae in battery cages stocked with laying hens. Veterinary Parasitology, 147(-4):289-296. Morand S., Sorcib G. 1988. Determinants of Life-history Evolution in Nematodes. Parasitology Today 14(5):193-196. Morand S, Simkova A, Matejusová I, Plaisance L, Verneau O et Desdevises Y. 2002. Investigating patterns may reveal processes: evolutionary ecology of ectoparasitic monogeneans. International Journal of Parasitology 32:111-119. Moss WW. 1967. Some new analytic and graphic approaches to numerical taxonomy, with an example from the Dermanyssidae (Acari). Systematic zoology 16:177-207. Moss WW. 1968. An Illustrated Key to the Species of the Acarine Genus Dermanyssus (Mesostigmata: Laelapoidea: Dermanyssidae). Journal of Medical Entomology 1:67-84. Moss WW. 1978. The mite genus Dermanyssus: a survey, with description of Dermanyssus trochilinis, n. sp., and a revised key to the species (Acari: Mesostigmata: Dermanyssidae). Journal of Medical Entomology 14(6):627-640. Moss WW, Mitchell CJ, Johnston DE. 1970. New North American host and distribution records for the mite genus Dermanyssus (Acari: Mesostigmata: Dermanyssidae). Journal of Medical Entomology 7:589-593. Moss, WW, Camin JH. 1970. Nest parasitism, productivity, and clutch size in Purple Martins. Science 168:1000-1003. Mullens BA, Hinkle NC, Robinson LJ, Szijj CE. 2001. Dispersal of Northern Fowl Mites, Ornithonyssus sylviarum, Among Hens in an Experimental Poultry House. Journal of applied poultry research 10: 60-64. Mullens B.A., Velten R.K., Hinkle N.C., Kuney D.R., Szijj C.E. 2004. Acaricide resistance in northern fowl mite (Ornithonyssus sylviarum) populations on caged layer operations in Southern California. Poultry Science 83:365-374 Nature Midi-Pyrénées (association). 2001. Etude et prospection du Martinet Pâle Apus pallidus en Midi- Pyrénées. http://www.premiumwanadoo.com/naturemp/6Proteger/Etudes/RapportMartinetPale2.pdf Navajas M, Fenton B. 2000. The application of molecular markers in the study of diversity in Acarology: a review. Experimental and Applied Acarology 24:751-774. Nei M. 1987. Molecular Evolutionary Genetics. Columbia University Press. Newberry K. 2008. The effects on domestic infestations of Cimex lectularius bedbugs of interspecific mating with C. hemipterus. Medical and Veterinary Entomology 3(4):407-414.
255 Nisbet AJ, Huntley JF, MacKellar A, Sparks N, McDevitt R. 2006. A house dust mite allergen homologue from poultry red mite Dermanyssus gallinae (De Geer). Parasite Immunology 28:401- 405. Nordenfors H, Höglund J, Uggla A. 1999. Effects of Temperature and humidity on oviposition, molting, and longevity of Dermanyssus gallinae (Acari : Dermanyssidae). Journal of Medical Entomology 36:68-72 Nordenfors H, Hoglund J. 2000. Long term dynamics of Dermanyssus gallinae in relation to mite control measures in aviary systems for layers. British Poultry Science 41(5):533-540 Nosek J, Lichard M. 1962. Beitrag zur Kenntnis der Vogelnestfauna. Entomologické problémy (Bratislava) 2:29-51 Oliver JH Jr. 1977. Cytogenetics of mites and ticks. Annual Review of Entomology 22:407-29 Oliver JH Jr. 1966. Notes on Reproductive Behavior in the Dermanyssidae. Journal of Medical Entomology 3:1:29-35 Pacejka AJ, Thompson CF. 1996. Does removal of old nests from nestboxes by researchers affect mite populations in subsequent nests of House Wrens? Journal of Field Ornithology 67:558-564. Pacejka J, Gratton CM, Thompson CF. 1998. Do potentially virulent mites affect house wren reproductive success? – Troglodytes aedon. Ecology 5:79. Page RDM, Hafner MS. 1996. Molecular phylogenies and host–parasite cospeciation: gophers and lice as a model system. In: P.H. Harvey, A.J. Leigh-Brown, J. Maynard Smith and S. Nee, Editors, New Uses for New Phylogenies, Oxford University Press, Oxford (1996), pp. 255–270. Page RD, Charleston MA. 1997. From gene to organismal phylogeny: reconciled trees and the gene tree/species tree problem. Molecular Phylogenetics and Evolution 7(2):231-40. Panchal M. 2007. The automation of Nested Clade Phylogeographic Analysis. Bioinformatics 23(4):509- 510. Panchal M, Beaumont MA. 2007. The automation and evaluation of nested clade phylogeographic analysis. Evolution 61(6):1466-80. Peterson, A. Version 8.07. 25/08/2007. ‘Zoonomen Nomenclatural data’ http://www.zoonomen.net/avtax/frame.html. Petit RJ. 2007. The coup de grâce for the nested clade phylogeographic analysis? Molecular Ecology 17(2):516-518. Phillis W. 1972. Seasonal abundance of Dermanyssus hirundinis and D. americanus (Mesostigmata: Dermanyssidae) in nests of the House Sparrow. Journal of Medical Entomology 9:111-112 Phillis WA, Cromroy HL, Denmark HA. 1976. New Host and distribution records for the mite genera Dermanyssus, Ornithonyssus and Pellonyssus (Acari: Mesostigmata: Laelapoidea) in Florida). The Florida Entomologist 59:1:89-92. Prasad RS. 1987. Nutrition and reproduction in haematophagous arthropods. Proceedings of the Indian Academy of Sciences (Animal Sciences) 96(3):253-273. Price PW. 1975. Insect ecology. J. Wiley, New York. Price RD, Hellenthal RA, Palma RL. 2003. World checklist of chewing lice with host associations and keys to families and genera. In: R.D. Price, R.A. Hellenthal, R.L. Palma, K.P. Johnson and D.H. Clayton, Editors, The Chewing Lice. World checklist and Biological Overview, Illinois Natural History Survey Special Publication 24:1–448. Radovsky F.J. 1966. Revision of the Macronyssid and Laelapid mites of bats: outline of classification with descriptions of new genera and new type species. Journal of Medical Entomology 3(1):93-99 Radovsky FJ. 1969. Adaptive radiation in parasitic Mesostigmata. Acarologia 11:450-483.
256 Radovsky FJ. 1994. The Evolution of Parasitism and the Distribution of Some Dermanyssoid Mites (Mesostigmata) on Vertebrate Hosts. In Mites Ecological and Evolutionary Analyses of Life- History Patterns, ed. MA Houck, USA. Reinhardt K, Siva-Jothy MT. 2007. Biology of the Bed Bugs (Cimicidae). Annual Review of Entomology 52:351-74. Rosen S, Yeruham I, Braverman Y. 2002. Dermatitis in humans associated with the mites Pyemotes tritici, Dermanyssus gallinae, Ornithonyssus bacoti and Androlaelaps casalis in Israel. Medical and Veterinary Entomology 16:4:442-444. Rosen S, Hadani A, Shoham D. 1985. Parasitic mites (Acarina, Arachnoidea) in wild birds trapped in poultry farms in Israel. Acarologia 26:79-85. Roy L, Guilbert E, Bourgoin T. 2007. Phylogenetic patterns of mimicry strategies in Darnini (Insecta: Hemiptera: Membracidae). Annales de la Société Entomologique de France, 43(3) : 273-288. Sahibi H, Rhalem A. 2007. Dermanyssus gallinae: Acarien particulièrement agressif et négligé à tort au Maroc. 1er Colloque International d'Acarologie et d'Entomologie Médicales, Rabat, 02-03 novembre 2007. Sokóá R, Szkamelski A, Barski D. 2008. Influence of light and darkness on the behaviour of Dermanyssus gallinae on layer farms. Pol J Vet Sci. 11(1):71-3. Sparagano O, Pavliüeviü A, Murano T, Camarda A, Sahibi H, Kilpinen O, Mul M, van Emous R, le Bouquin S, Hoel K, Cafiero M. 2009. Prevalence and key figures for the poultry red mite. Dermanyssus gallinae infections in poultry farm systems. Experimental and Applied Acarology 48:1:3-10 Strandtmann RW, Wharton GW. 1958. A manual of mesostigmatid mites parasitic on vertebrates Institute of Acarology, Contribution No. 4, University of Maryland, College Park, Maryland. Szalanski AL, Austin JW, McKern JA, Steelman CD, Miller DM, Gold RE. 2006. Isolation and Characterization of Human DNA From Bed Bug, Cimex lectularius L. (Hemiptera : Cimicidae) Blood meals. Journal of Agricultural and Urban Entomology 23(3):189-194. Szalanski AL, Austin JW, McKern JA, Steelman CD, Gold RE. 2008. Mitochondrial and Ribosomal Internal Transcribed Spacer 1 Diversity of Cimex lectularius (Hemiptera: Cimicidae). Journal of Medical Entomology 45(2):229-236. Templeton AR. 1998. Nested clade analyses of phylogeographic data: testing hypotheses about gene flow and population history. Molecular Ecology 7:399-412. Templeton AR, Routman E, Phillips CA. 1995. Separating population structure from population history: a cladistic analysis of the geographical distribution of mitochondrial-DNA haplotype in the Tiger Salamander Ambystoma tigrinum. Genetics 140:767-782. Todisco G, Paoletti B, Giammarino A, Manera M, Sparagano OA, Iorio R, Giannella B, Robbe D. 2008. Comparing therapeutic efficacy between ivermectin, selamectin, and moxidectin in canaries during natural infection with Dermanyssus gallinae. Annals of the New York Academy of Sciences 1149:365-7. Troughton DR, Levin LR. 2007. Life Cycles of Seven Ixodid Tick Species (Acari: Ixodida) under Standardized Laboratory Conditions. Journal of Medical entomology 44(5):732-740. Tuovinen T. 2008. Predatory arthropods as biocontrol agents of Dermanyssus gallinae? In BSP Spring, Trypanosomiasis/Leishmaniasis & Malaria Meetings.March 30th-April 2nd 2008. (poster). Uchikawa K, Takahashi M. 1985. Contribution to mites of the genus Dermanyssus De Geer distributed in Japan (Acarina, Mesostigmata) Japanese Journal of Sanitary Zoology 36:3:233-237. Valera F, Casas-Crivillé, Hoi H. 2003. Interspecific Parasite exchange in a mixed colony of birds. Journal of Parasitology 89(2):245-250.
257 Valiente Moro C, Chauve C, Zenner L. 2005. Vectorial role of some dermanyssoid mites (Acari, Mesostigmata, Dermanyssoidea). Parasite 12(2): 99-109. Valiente Moro, C., Fravalo, P., Amelot, M., Chauve, C., Zenner, L., Salvat, G. 2007. Colonization and invasion in chicks experimentally infected with Dermanyssus gallinae contaminated by Salmonella enteritidis. Avian Pathol. 36(4): 307-311. Valiente Moro C., De Luna C.J., Tod A., Guy J.H., Sparagano O.A.E., Zenner L. 2009. The poultry red mite (Dermanyssus gallinae): a potential vector of pathogenic agents. Experimental and Applied Acarology 48 :93-104. Verneau O, Bentz S, Sinappah ND, du Preez L., Whittington I, Combes C. 2002. A View of early vertebrate evolution inferred from the phylogeny of polystome parasites (Monogenea: Polystomatidae). Proceedings of the Royal Society of London 269:535-543. Vitzthum H.G. 1921.Acarologische Beobachtungen. 4 Riehe. Archiv für Naturgeschichte 86A(10):1-69 Williamson M. 1996. Biological invasions. New York : Chapman & Hall. Witalinski W. 2000. Aclerogamasus stenocornis sp. n., a fossil mite from the Baltic amber (Acari: Gamasida: Parasitidae). Genus 11:619-662. Wood HP. 1917. The chicken mite: its life history and habits. United States Department of Agriculture, Washington, DC Bulletin 553:1-14. Wright HW, Bartley K, Nisbet AJ, McDevitt RM, Sparks NHC, Brocklehurst S, Huntley JF. 2009. The testing of antibodies raised against poultry red mite antigens in an in vitro feeding assay; preliminary screen for vaccine candidates. Experimental and Applied Acarology 48:81-91. Zeman P, Jurík M. 1981. A contribution to the knowledge of fauna and ecology of gamasoid mites in cavity nests of birds in Czechoslovakia. Folia Parasitologica (Prague) 28: 265-271. Zemskaya AA. 1968. [Sparrow mite Dermanyssus passerinus Berlese and Trouessart.] Medicinskaâ parazitologiâ i parazitarnye bolezni 37:313-319 (en Russe). Zemskaya AA. 1971. [Mites of the family Dermanyssidae Kolenati, 1859, of the U.S.S.R. fauna]. Medicinskaâ parazitologiâ i parazitarnye bolezni 40: 709-717 (en Russe). Zemskaya AA et Ilienko AI. 1958. [Blood-feeding mites on house and field sparrows in Moscow and environs.] Medicinskaâ parazitologiâ i parazitarnye bolezni 27:475-481 (en Russe).
258 10 Annexes
10.1 Annexe 1 : apercu de la classification des Mesostigmata selon Hallan 2005 Famille Discourellidae Famille Uropodidae Ordre Famille Metagynuridae Mesostigmata Famille Oplitidae Famille Trachyuropodidae Sous-ordre Diarthrophallina Sous-ordre Heatherellina Super-Famille Diarthrophalloidea Famille Heatherellidae Famille Diarthrophallidae Sous-ordre Sejina Sous-ordre Cercomegistina Super-Famille Sejoidea Super-Famille Cercomegistoidea Famille Sejidae Famille Cercomegistidae Sous-ordre Arctacarina Famille Saltiseiidae Super-Famille Arctacaroidea Famille Asternoseiidae Famille Arctacaridae Famille Davacaridae Sous-ordre Microgyniina Famille Seiodidae Famille Pyrosejidae Super-Famille Microgynioidea Famille Nothogynidae Sous-ordre Antennophorina Famille Microgyniidae Super-Famille Aenicteguoidea Sous-ordre Epicriina Famille Aenicteguidae Famille Messoracaridae Super-Famille Epicrioidea Famille Physalozerconidae Famille Epicriidae Famille Ptochacaridae Famille Dwigubskyiidae Super-Famille Antennophoroidea Famille Coprozerconidae Famille Antennophoridae Famille Zerconidae Super-Famille Celaenopsoidea Sous-ordre Uropodina Famille Neotenogyniidae Super-Famille Uropodoidea Famille Celaenopsidae Famille Protodinychidae Famille Costacaridae Famille Thinozerconidae Famille Schizogyniidae Famille Polyaspididae Famille Megacelaenopsidae Famille Trachytidae Famille Triplogyniidae Famille Dithinozerconidae Famille Meinertulidae Famille Nenteriidae Famille Diplogyniidae Famille Trematuridae Famille Euzerconidae Famille Macrodinychidae Super-Famille Fedrizzioidea Famille Trigonuropodidae Famille Fedrizziidae Famille Urodinychidae Famille Klinckowstroemiidae Famille Dinychidae Famille Promegistidae Famille Uroactinidae Famille Paramegistidae Famille Circocyllibamidae Super-Famille Megisthanoidea Famille Deraiophoridae Famille Hoplomegistidae
259 Famille Megisthanidae Famille Halolaelapidae Super-Famille Parantennuloidea Famille Ameroseiidae Famille Parantennulidae Famille Podocinidae Famille Philodanidae Super-Famille Dermanyssoidea Sous-ordre Parasitina Famille Trichoaspididae Famille Larvamimidae Super-Famille Parasitoidea Famille Leptolaelapidae Famille Parasitidae Famille Varroidae Sous-ordre Dermanyssina Famille Laelapidae Super-Famille Rhodacaroidea Famille Haemogamasidae Famille RhodacaroideaIncertae Famille Pneumophionyssidae Famille Ologamasidae Famille Dermanyssidae Famille Euryparasitidae Famille Hirstionyssidae Famille Rhodacaridae Famille Hystrichonyssidae Famille Digamasellidae Famille Macronyssidae Famille Laelaptonyssidae Famille Rhinonyssidae Famille Panteniphididae Famille Spinturnicidae Super-Famille Veigaioidea Famille Spelaeorhynchidae Famille Veigaiidae Famille Halarachnidae Famille Raillietiidae Super-Famille Eviphidoidea Famille Entonyssidae Famille Macrochelidae Famille Ixodorhynchidae Famille Parholaspididae Famille Omentolaelapidae Famille Pachylaelapidae Famille Dasyponyssidae Famille Megalolaelapidae Famille Manitherionyssidae Famille Eviphididae Super-Famille Ascoidea Sous-ordre Heterozerconina Famille Ascidae Famille Heterozerconidae Famille Phytoseiidae Famille Discozerconidae Famille Otopheidomenidae
260 Se former / CONDUITE À TENIR /
DERMANYSSUS GALLINAE, ECTOPARASITE DES VOLAILLES Pou rouge : diagnostic et lutte contre l’infestation Dermanyssus gallinae est un parasite redouté en élevage de pondeuses. Cet acarien discret est un vecteur expérimental de salmonelles. Les traitements sont limités par l’absence de médicament avec une LMR définie dans les œufs.
ermanyssus gallinae (the red fowl Les étapes essentielles mite, actuellement dans la famille des Dermanyssidés, voir le TABLEAU complémentaire “Place de D. Étape 1 : identifier les gallinae parmi les acariens” sur contextes d’infestation DPlanète-vet) est un acarien hématophage • Chez divers oiseaux parasite des oiseaux et résistant au jeûne. La (peuvent aussi piquer répartition des poux rouges est particulière à des mammifères chaque élevage. Elle doit donc être clairement en l’absence d’oiseaux). identifiée afin de mettre en place des moyens • Surtout chez les poules de lutte efficaces. Le pou rouge ne doit pas être pondeuses. Présent aussi confondu avec d’autres espèces rencontrées en en élevage amateur. élevage de volailles en France. • Se cache dans des abris variés de l’environnement. • Baisse de production, déclassement des œufs, Première étape : voire mortalité. Rôle vectoriel. identifier les contextes d’infestation Étape 2 : reconnaître le parasite 1. Aire de répartition • Diagnose en élevage de D. gallinae est présent en Europe, alors qu’Orni- Cliché : L. Roy volailles en France, thonyssus sylviarum, une espèce apparentée, PHOTO 1. Agrégat de D. gallinae fraîchement à l’examen microscopique gorgés sous un amas de fientes séchées, sévit plutôt en Europe du Nord et en Amérique accumulées sur le seuil d’une trappe de sortie x 40 à x 400 de femelles du Nord. Cette dernière espèce a été signalée adultes ébouillantées : (élevage de poules pondeuses en plein air). examiner les chélicères, en France il y a quelques années [3], mais n’a pas la couleur, pour pas été retrouvée depuis. au sol chez les pondeuses. L’acarien est distinguer D. gallinae également rencontré dans les élevages des autres acariens. 2. Espèces atteintes “amateurs”, même si les infestations massives Le pou rouge est capable de parasiter un grand sont rares. En effet, les conditions sont favora- nombre d’oiseaux (plus de trente espèces bles à l’hébergement d’arthropodes prédateurs Étape 3 : traiter recensées). En l’absence de volatiles, il peut des poux et la densité des volailles au mètre • L’animal : non autorisé aussi piquer des mammifères, notamment les carré est moindre. et insuffisant. chevaux et des rongeurs, ainsi que l’homme. Il • Son environnement : peut alors provoquer une gêne chez le person- 4. Facteurs favorisant le maintien du pou autorisé pendant le vide nel, liée à des irritations cutanées et à une Les poux rouges se logent dans des abris variés sanitaire en pondeuse, éventuelle allergie à l’acarien. Le problème et souvent difficiles d’accès, notamment pour les nettoyage, puis majeur posé par D. gallinae ne se situe toutefois substances acaricides éventuellement utilisées : organophosphorés, certaines pyréthrinoïdes, pas dans les maisons de particuliers comme sous des fientes sèches, dans des amas de plumes avec beaucoup d’eau. cela a été suggéré, mais en aviculture [2]. ou de duvet, dans les fissures des murs, dans les En présence des volailles, interstices situés entre divers constituants des aucune LMR pour les œufs. 3. Types d’élevages infestés structures d’élevage, notamment les petits À peu près maîtrisé dans les élevages de volailles éléments métalliques ou en matière plastique de chair, parfois présent chez les reproducteurs, qui servent à unir les barreaux des cages en par Lise Roy*, le pou rouge pose surtout un problème en batterie, des perchoirs, pondoirs et/ou caillebo- élevage de poules pondeuses car la bande est tis au sol, etc. (PHOTO 1). La distribution de Claire Valiente Moro* maintenue en production plus longtemps et les l’acarien est en outre variable d’un élevage, voire et Claude Chauve* traitements sont limités par les directives d’un bâtiment à l’autre. En l’absence de tout hôte, * Laboratoire de parasitologie européennes sur les limites maximales de D. gallinae peut survivre plusieurs mois. Sa de l’ENV Lyon, UMR Inra-ENVL 958, résidus (LMR) dans les œufs. Il trouve des résistance au jeûne est fonction de son stade de 1, avenue Bourgelat, conditions de développement optimales dans développement, mais aussi de la température et 69280 Marcy-l’Étoile les élevages avicoles actuels : en cage comme de l’hygrométrie relative. Les protonymphes et 46 Le Point Vétérinaire / N° 266 / Juin 2006 / les deutonymphes, les mâles adultes et les femelles non encore gorgées peuvent survivre Cycle parasitaire de D. gallinae sans nourriture pendant plusieurs mois (huit ou neuf mois pour les deutonymphes selon certains auteurs). La prise du premier repas de sang nécessaire à la maturation des œufs raccourcit la longévité des femelles. Toutefois, celles qui se nourrissent et pondent sans interruption n(1) repas semblent vivre plus longtemps que celles qui, de sang Œuf Larve après un premier cycle gonotrophique (voir la FIGURE “Cycle parasitaire de D. gallinae”), se trouvent privées de nourriture et ne peuvent pas Femelle enchaîner un deuxième cycle. C A 5. Impact en élevage de volailles Cet ectoparasite qui ponctionne le sang des poules peut provoquer des pertes économiques Adulte N1 non négligeables. Source de stress et d’irrita- Mâle tions cutanées, il perturbe les oiseaux. Il B engendre du picage et une détérioration du D plumage, une augmentation de la consomma- tion de nourriture par les volailles, accompa- gnés parfois d’une chute de la ponte. Les œufs N2 1 repas sont déclassés en raison des taches de sang dues 1 repas de sang à des poux écrasés sur la coquille (jusqu’à 5 % de sang d’œufs déclassés sur l’ensemble de la produc- tion d’une bande). Lors d’infestation massive, qui survient rapidement en l’absence d’inter-
vention, le nombre d’érythrocytes et la concen- Clichés : L. Roy tration en hémoglobine peuvent diminuer. Dans les conditions optimales, le cycle s’accomplit en une à deux semaines. Les larves L’interprétation de cette anémie n’est cependant (PHOTO A) ne se nourrissent jamais. Environ 24 heures après leur éclosion, elles muent pas facile et les variations dans la composition en protonymphes (PHOTO B : encore à jeun). Les protonymphes (N1) et deutonymphes sanguine demeurent souvent peu significatives. (N2) ont besoin d’un repas de sang pour accomplir leur métamorphose (environ 24 heures). La femelle adulte a besoin d’un repas de sang avant chaque ponte, La mortalité par exsanguination constatée au 12 heures à 24 heures après chaque repas. Elle pond jusqu’à sept œufs à la fois laboratoire est difficile à évaluer sur le terrain. et peut accomplir jusqu’à huit de ces cycles gonotrophiques dans sa vie, parfois Comme les tiques, cet acarien est en outre avec une seule fécondation. Plus la femelle a réalisé de pontes, plus elle pond d’œufs susceptible de transmettre des maladies entre à la fois. (PHOTO C : femelle adulte multipare en cours de digestion). volailles en raison de son comportement Le mâle adulte (PHOTO D : état de digestion plus avancé) ne semble pas avoir besoin de se nourrir. Il semble rechercher les deutonymphes fraîchement gorgées, et donc en hématophage (voir l’ENCADRÉ “Rôle vectoriel passe d’accomplir leur mue imaginale, et s’accouple avec elles peu après l’exuviation. potentiel du pou rouge des volailles”). (1) n = nombre d’ovipositions.
Rôle vectoriel potentiel du pou rouge des volailles
! Au même titre que les tiques, l’implication de cet dans les matières fécales), sans en être pour autant acarien hématophage dans la transmission vectorielle un réservoir naturel. Il serait vecteur potentiel et de plusieurs maladies a été suspectée (voir l’ENCADRÉ réservoir avéré d’Erysipelothrix rhusiopathiae, complémentaire “Anthropodes vecteurs définitions”, bactérie responsable du rouget du porc (une sur Planete-vet). Les travaux sont encore peu zoonose). Il a aussi été expérimentalement nombreux sur ce sujet et, lorsqu’ils existent, ils sont démontré que D. gallinae peut être contaminé en souvent limités. Le rôle du parasite est donc certaine- se nourrissant sur des animaux infectés par Coxiella ment sous-estimé dans l’apparition, le maintien et la burnetii, agent de la fièvre Q (une zoonose). Actuel- propagation des certaines infections. lement, dans la superfamille des Dermanyssoidea, ! Pour les salmonelles (zoonose), le rôle de vecteur seule Liponyssoides sanguineus, espèce apparte- biologique expérimental du pou rouge a été nant aux Dermanyssidés, est reconnue comme le récemment démontré par notre équipe. Des vecteur principal d’un agent pathogène (Rickettsia transmissions transovarienne et transstadiale des akari). salmonelles ont en outre été observées. Si le rôle ! Des virus animaux, dont ceux de la variole aviaire de vecteur naturel était confirmé par l’isolement de et de la maladie de Newcastle, sont parfois trouvés salmonelles à partir de prélèvements du terrain, associés à D. gallinae. Ainsi, le rôle de vecteur D. gallinae pourrait être un facteur favorisant la biologique a été établi pour les virus des encépha- persistance des infections à salmonelles pendant lites équines de l’Est et de l’Ouest. Un rôle de le vide sanitaire. Des bactéries ont été isolées à vecteur mécanique vis-à-vis du virus de l’encépha- partir d’acariens prélevés sur le terrain, mais un plus lite équine vénézuélienne a également été associé. grand nombre de cas serait nécessaire pour Les virus de l’encéphalite de Saint-Louis et des conclure au rôle de vecteur naturel de D. gallinae. encéphalites à tiques ont été isolés sur ces acariens, ! Les genres Listeria et Pasteurella, autres agents sans que leur rôle vecteur ou celui de réservoir de zoonose, ont été isolés sur des poux rouges, aient été démontrés. Ces virus équins sévissent en mais le rôle vecteur n’a pas encore été étudié pour Amérique, mais D. gallinae pourrait porter des virus ces bactéries. D. gallinae serait un vecteur occasion- présents en Europe, comme celui de la maladie de nel de spirochètes (élimination de ces bactéries Marek qui affecte les volailles. !!
/ N° 266 / Juin 2006 / Le Point Vétérinaire 47 Se former / CONDUITE À TENIR /
!! est possible en élevage de volailles en France (voir Deuxième étape : l’ENCADRÉ “Définitions” et l’ENCADRÉ complémen- reconnaître le parasite taire “Distinction entre le mâle et la femelle” sur Planète-vet ). Une grille de diagnose peut être 1. Modalités pratiques de diagnose utilisée avec des acariens de taille moyenne à simplifiée grande (au minimum 0,5 à 1 mm de long) (voir L’examen microscopique de femelles adultes aux l’ENCADRÉ “Diagnose de D. gallinae”). En deçà, le grossissements x 40 à x 400, en considérant en parasite appartient à un autre groupe ou bien il particulier les chélicères, permet de faire la distinc- s’agit d’un stade inapproprié à la diagnose. Il tion entre les différents acariens dont la présence convient d’examiner plusieurs individus.
Diagnose de D. gallinae
Acarien femelle adulte de 0,5 à 1 mm au microscope optique, grossissements x 40 à x 400, issu d’un élevage de volailles en France
! Coxae (= bases des pattes) distinctes ! Coxae réduites à d'étroites plaques formant chacune 1 article à part entière chitineuses sillonnant la cuticule ventrale ! Coxae plus ou moins contiguës, ! Généralement, coxae en deux groupes distincts : occupant les deux tiers antérieurs du corps paires I et II situées vers l’avant, nettement séparées des paires III et IV, situées à l'arrière ien dessi xa b née Co I et II Il ne s’agit pas d'un parasitiforme
III et IV
! Articles des chélicères ! Articles des chélicères porteurs des pinces porteurs des pinces ( ) aussi plus minces que les épais ( ), voire plus pédipalpes et très allongés épais que les pédipalpes (dépassant souvent ( ) et ne dépassant pas de beaucoup la longueur des pédipalpes L’acarien appartient probablement la longueur de ceux-ci) ! Pinces très nettes, dentées à l'une des familles suivantes : ! Pinces réduites Laelapidé, Haemogamasidé, Parasitidé, Macrochélidé.
! Pinces atrophiées, non visibles au microscope optique Il s’agit d’un Dermanyssidé, et très probablement ! Corps dur ! Corps mou de D. gallinae (se brisant aux contours à la pression) peu définis aux contours ! Pinces visibles arrondis très nets ou non ! Pinces visibles, fines et non dentées ! Pinces visibles
Il s’agit d’un Macronyssidé
Il s’agit d’un Uropodoidea
48 Le Point Vétérinaire / N° 266 / Juin 2006 / La couleur du parasite ne présente aucun intérêt Adaptation des chélicères à l’hématophagie pour la diagnose, car la cuticule de nombreux acariens, dont celle de D. gallinae, est transpa- ! À gauche, les chélicères d’une femelle appartenant à une espèce non parasite apparen- rente et laisse apparaître les organes et les tée aux Dermanyssidés donnent une idée de l’état ancestral, avec ses chelae ou pinces liquides internes. L’acarien n’est donc rouge que nettement développées (microscope optique). lorsqu’il vient de prendre un repas de sang. ! Au centre, les chélicères d’une femelle de D. gallinae se présentent comme de fins Avant l’examen, il convient de tuer les acariens cheveux au microscope optique. dans de l’eau bouillante pour favoriser le ! À droite, une vue ventrale au microscope électronique à balayage des pièces buccales déploiement des chélicères (souvent rétractées, d’une femelle de D. gallinae permet de mieux comprendre leur conformation : les chélicè- surtout chez les Macronyssidés et les Dermanys- res amincies et modifiées en gouttières peuvent se réunir pour former un sorte de tuyau. sidés). Une décoloration avec de la potasse à À leur base, un fourreau permet leur rétraction. 10 % (chauffée au bain-marie pendant quinze minutes) ou à l’acide lactique (à température ambiante, pendant un à trois jours) facilite l’observation. Pour rendre la décoloration plus efficace, il est possible de percer préalablement chaque individu vers l’arrière du corps à l’aide d’une aiguille fine, à la loupe binoculaire. Schématiquement, hormis dans les familles des Macronyssidés, des Dermanyssidés et dans la superfamille des Uropodoidea, les chélicères des femelles adultes sont massives (articles épais, non Clichés : L. Roy filiformes) et munies de pinces, ou chelae, nettes et extrêmement chitinisées (forme “ancestrale”, (voir l’ENCADRÉ “Adaptation des chélicères à Définitions l’hématophagie”). Les femelles adultes des Macronyssidés, hématophages obligatoires, ainsi ! Parasitiforme : l’une des deux grandes divisions que des Uropodoidea, présentent des chélicères du groupe des acariens, englobant entre autres les fortement allongées et aux chelae, réduites mais Mésostigmates tels D. gallinae, Ornithonyssus spp., nettement dessinées. Chez les Dermanyssidés, Varroa spp. et les Métastigmates ou Ixodida, c’est- à-dire les tiques. L’autre division est celle des hématophages obligatoires, l’allongement est acariformes. également marqué et les chelae, sont atrophiées, ! Coxa : article basal des pattes. Les pattes des indistinctes au microscope optique, même à fort acariens sont constituées de six articles : coxa, grossissement (chélicères des femelles adultes trochanter, fémur, genou, tibia, tarse (depuis la base filiformes, semblables à des cheveux). jusqu’à l’extrémité). Ces éléments de diagnose succincts suffisent à ! Chélicères : appendices buccaux pairs des infirmer ou à confirmer une infestation à acariens, constitués de deux articles, le second D. gallinae dans un élevage de poules pondeu- portant une pince ou chela. Rétractiles chez certai- ses en France, mais en aucun cas pour les autres nes espèces. espèces d’acariens. Quelque 700 espèces sont ! Pédipalpes : appendices buccaux pairs, sembla- incluses dans la famille des Laelapidés, plus de bles à de courtes pattes et encadrant les chélicères. 100 dans celle des Macronyssidés, une soixan- taine chez les Haemogamasidés… Les familles d’acariens potentiellement présentes en élevage bles à ceux engendrés par D. gallinae. Ainsi, ne sont pas toutes citées ici. Ornithonyssus bacoti (Macronyssidé) est un ectoparasite hématophage inféodé aux rongeurs. 2. Risque de confusion avec d’autres Androlaelaps casalis (Laelapidé) est un prédateur acariens d’autres arthropodes et un parasite hémato- Dans les élevages de pondeuses en France, phage seulement occasionnel (signalé en D. gallinae est pratiquement la seule espèce colonies importantes dans un élevage de dindes infestante. La distinction entre D. gallinae et les auquel il n’infligeait aucun dommage). Des autres espèces du genre Dermanyssus Haemogamasidés, dont certaines espèces parasi- (D. hirundinis, D. gallinoides, etc.) est difficile. tent les rongeurs, peuvent aussi être rencontrés Cependant, ce sont plutôt des parasites des dans les élevages. Des espèces de la superfamille oiseaux sauvages, rarement observés dans les des Uropodoidea, saprophages et/ou prédatri- élevages (et systématiquement associés à des ces, sont fréquemment présentes dans les quantités massives de D. gallinae). Certains élevages au sol, ainsi que chez les éleveurs parasites de rongeurs du genre Liponyssoides “amateurs”. Leiodinychus krameri (Dinychidé), Un TABLEAU spp., ressemblant au pou rouge et appartenant couramment rencontré dans la poussière des complémentaire aussi à la famille des Dermanyssidés, peuvent greniers à foin, est parfois présent dans les “Place de D. gallinae aussi être rencontrés, mais ils sont rares. poulaillers. Signalé comme un parasite parmi les acariens” Les autres espèces d’acariens parasites potentiel- occasionnel (par Neveu-Lemaire en 1938), il et un ENCADRÉ lement présents dans les élevages appartiennent n’est généralement à l’origine d’aucun problème. complémentaire à d’autres familles. Les caractères discriminants De nombreuses espèces de Laelapidés sont “Distinction entre le mâle sont donc plus accessibles. Ils sont soit parasi- exclusivement prédatrices et incapables de et la femelle” tes d’autres espèces que les volailles (souris parasiter la volaille (Hypoaspis spp., par sont consultables notamment, potentiellement présentes dans les exemple). D’autres familles de prédateurs sont sur le site bâtiments), soit parasites non obligatoires et souvent présentes, par exemple des Macroche- www.planete-vet.com incapables de provoquer des dégâts compara- lidés, des Parasitidés. Les espèces de cette Rubrique bibliographie !! / N° 266 / Juin 2006 / Le Point Vétérinaire 49 Se former / CONDUITE À TENIR /
!! dernière famille, contrairement à ce que semble les bactéries, les champignons, etc. L’efficacité de (1) Index phytosanitaire de l’association de coordination indiquer leur nom, ne sont pas parasites ou tous ces produits est difficile à évaluer car ils n’ont technique agricole Acta 2006. seulement parasites occasionnels. pas fait l’objet d’une étude précise et officielle, et les avis en provenance du terrain sont dispara- tes, parfois contradictoires. Certaines différences Bibliographie Troisième étape : traiter d’efficacité constatées par les éleveurs et les autres acteurs de la filière sont parfois dues à des 1 - Beugnet F, Chauve C, Gauthey D. gallinae M., et coll. Resistance of the red La lutte contre se heurte à deux variations dans les protocoles d’application des poultry mite to pyrethroids in obstacles majeurs : sa biologie particulière et produits. La rigueur du protocole d’application France. Vet. Rec. 1997;140:577- les limitations réglementaires. et une connaissance exhaustive de la répartition 579. du parasite au sein de l’élevage semblent être les 2 - Bertrand, M. Note d’information 1. Traiter l’animal n’est pas suffisant clés d’une lutte efficace. D. gallinae s’abrite souvent sur un espèce particulièrement agressive d’Acarien : Dermanyssus Les traitements appliqués directement sur les en petits groupes dans des interstices particuliè- gallinae (DeGeer, 1778). Insectes. poules sont superflus car D. gallinae est un rement étroits, à tel point qu’il est parfois difficile 1998;111:21-23. ectoparasite nidicole. À la différence de O. sylvia- d’envisager que des poux rouges peuvent s’y loger. 3 - Bruneau, A., Dernburg, A., rum, le pou rouge ne séjourne pas longtemps sur Il convient donc de ne pas se limiter à traiter les Chauve, C. et coll. First report of the northern fowl mite son hôte. Le parasite ne grimpe sur l’oiseau que zones où sont réunis les agrégats les plus visibles. Ornithonyssus sylviarum in France. pour prendre un repas de sang, principalement Vet. Rec. 2002;150:413-414. la nuit, pendant une demi-heure à une heure et 3. Émergence d’une chimiorésistance ? 4 - Chauve C. The poultry red mite demie. Une fois le sang prélevé, il retourne dans Les éleveurs se heurtent parfois à une diminu- Dermanyssus gallinae (De Geer, 1778): current situation and future un abri : la litière, les anfractuosités des murs, tion de l’efficacité des produits utilisés, qui peut prospects for control. Vet. Parasitol. etc. Ce sont les lieux où il séjourne qu’il faut être due à un défaut d’application. L’apparition 1998;79:239-245. traiter pour le détruire. La lutte doit être instau- de phénomènes de résistance à certaines molécu- 5 - Cosoroaba I. Observation rée précocement car le cycle du pou rouge peut les acaricides est aussi possible. Au laboratoire, d’invasions massives par être extrêmement rapide si un hôte est disponi- des différences de sensibilité à certaines molécu- Dermanyssus gallinae (De Geer 1778), chez les poules élevées en ble et que la température et l’hygrométrie sont les acaricides ont été mises en évidence, aux batterie en Roumanie. Revue Méd. adéquates, comme c’est le cas dans les élevages pyréthrinoïdes notamment [1]. Des résistances Vét. 2001;152:1:89-96. de poules pondeuses. Un œuf du pou rouge peut marquées contre le DDT avait aussi été suspec- 6 - Evans GO et Till WM. Studies on alors évoluer pour devenir une femelle prête à tées. Elles semblent de faible intensité vis-à-vis the British Dermanyssidae (Acari: Mesostigmata). Part I. External pondre en une à deux semaines seulement. de quelques organophosphorés [14], sur des morphology. Bull. Br. Mus. (Nat. populations de poux rouges provenant d’éleva- Hist.) Zool. 1965; 13:249-294 2. Substances utilisables dans ges soumis à des pressions insecticides variables. 7 - Kirkwood A. Longevity of the l’environnement Les travaux sur ce sujet demeurent peu nombreux Mites Dermanyssus gallinae and et, en l’absence de souche sensible de référence, Liponyssus sylviarum. Exp. ! Parasitol. 1963;14:358-366. Pendant le vide sanitaire aucune véritable résistance n’a pu être démontrée. 8 - Moss WW. The mite genus Plusieurs acaricides au sens large peuvent être Dermanyssus : a survey, with appliqués lors du vide sanitaire (molécules 4. Moyens complémentaires de lutte description of Dermanyssus chimiques, mais aussi silice, extraits de plantes, Un programme lumineux par cycles courts trochilinis, n. sp., and a revised key to the species (Acari: etc.). L’utilisation des acaricides de synthèse en (quatre heures de lumière/deux heures d’obscu- Mesostigmata: Dermanyssidae). J. élevage de pondeuses n’est autorisée qu’entre rité) aide à lutter contre D. gallinae. La prolifé- Med. Entomol. 1978;14(6):627- deux bandes. Les principales molécules chimi- ration des acariens est probablement limitée 640 ques indiquées dans ce cadre sont des organo- par la perturbation de leurs repas. D’autres 9 - Neveu-Lemaire. Traité ® ® d’entomologie médicale et phosphorés (Actogard , Alfacron ). D’autres moyens complémentaires de lutte pourraient vétérinaire. Ed. Vigot, Paris. molécules destinées à lutter contre les arthro- être développés : phéromones répulsives, 1938;1339pp. podes (certains pyréthrinoïdes préconisés contre auxiliaires de lutte (acariens prédateurs, micro- 10 - Nordenfors H. Epidemiology les mouches(1)) sont aussi applicables. Il convient organismes entomopathogènes), etc. Ces voies and control of the poultry red mite, d’associer à ces traitements un dépoussiérage de recherche n’ont cependant pas encore abouti Dermanyssus gallinae. Thèse de doctorat, Swedish University of et un nettoyage efficace et minutieux afin à des applications sur le terrain. Agricultural Sciences, Uppsala. d’éliminer le maximum d’acariens. Diluer les 2000;43pp.Vappendix. matières actives dans de grandes quantités d’eau 11 - Reynaud MC, Chauve C, à pulvériser en augmente l’efficacité. Il est ainsi Il reste aussi beaucoup d’études à mener pour Beugnet F. Dermanyssus gallinae (De Geer, 1778) : reproduction recommandé d’imbiber au maximum les mieux appréhender le rôle vecteur de ce parasite, expérimentale du cycle et essai de structures. et notamment vis-à-vis de certains agents traitement par la moxidectine et pathogènes impliqués en santé publique. En effet, l’ivermectine. Rev. Méd. Vét. ! En présence de volailles Toulouse. 1997;148 :433-438. la résistance au jeûne du pou rouge des volailles, 12 - Valiente-Moro C, Chauve C et En présence de volailles, aucun médicament son comportement nidicole, son éventuelle Zenner L. Vectorial Role of Some stricto sensu n’est autorisé car il n’existe pas à ce chimiorésistance, ainsi que sa répartition ubiqui- Dermanyssoid Mites (Acari, jour de spécialité avec AMM disposant d’une taire en font une source potentielle de dissémi- Mesostigmata, Dermanyssoidea). Parasite 2005;12:99-109. LMR déterminée pour les œufs. Toutefois, nation d’agents pathogènes qui pourrait favori- d’autres produits qui n’ont pas le statut de médica- ser le maintien de zones d’endémie. ■ 13 - Wood HP. The chicken mite: its life history and habits. United ment, à base de soufre, de silice, d’huiles essentiel- States Department of Agriculture, les et/ou de pyrèthre naturel, sont employés, Congrès et internet Washington, DC, Bull. 1917;553: principalement en élevage de plein air. Certains 1-14. biocides utilisables en présence de poules peuvent - Lubac S, Dernburg A, Bon G et coll. Problématique et 14 - Zeman P, Z elezny J. The pratique d’élevage en poules pondeuses dans le sud-est Susceptibility of the Poultry Red être légalement appliqués. Il s’agit de produits de de la France contre les nuisibles : poux rouges et Mite, Dermanyssus gallinae (De traitement des bâtiments et des structures, et non mouches. Proc. 5es Journées de la recherche avicole, Geer, 1778), to some acaricides des poules, généralement à large spectre d’action, Tours, 26-27 mars 2003:101-104. under laboratory conditions. Exp. - http://insects.tamu.edu/research/collection/hallan/ Appl. Acarol. 1985;1:17-22. qui ne détruisent pas uniquement le pou rouge, mais aussi d’autres agents pathogènes tels que acari/0ReportHi.htm 50 Le Point Vétérinaire / N° 266 / Juin 2006 / ERRATA
La durée de prescription du carprofène est limitée par l’AMM
• Une erreur s’est glissée dans l’article “Gestion médicamenteuse de la douleur cancéreuse”de Roxane Steux, paru dans le n° 266 du Point Vétérinaire, page 56 : le carprofène ne possède pas d’AMM sans limitation de durée de prescription, cette dernière étant limitée à cinq jours dans le RCP. Il fallait donc lire : « - choisir un AINS avec AMM sans limitation de durée de prescription, possédant une bonne tolérance gastrique. L’AINS est alors administré en continu sur de longues périodes (exemple : méloxicam) ».
Figure “Diagnose de D. gallinae”
• Dans l’article “Pou rouge : diagnostic et lutte contre l’infestation” de Lise Roy et coll., publié dans le n° 266 du Point Vétéri- naire, la figure “Diagnose de D. gallinae” contient deux erreurs. Nous reproduisons donc cette figure corrigée. Diagnose de D. gallinae
84 Le Point Vétérinaire / N° 267 / Juillet-août 2006 / Exp Appl Acarol (2009) 48:19–30 DOI 10.1007/s10493-009-9249-z
Exploration of the susceptibility of AChE from the poultry red mite Dermanyssus gallinae (Acari: Mesostigmata) to organophosphates in field isolates from France
Lise Roy Æ Claude Chauve Æ Jean Delaporte Æ Gilbert Inizan Æ Thierry Buronfosse
Received: 10 October 2008 / Accepted: 23 January 2009 / Published online: 13 February 2009 Springer Science+Business Media B.V. 2009
Abstract The red fowl mite Dermanyssus gallinae (De Geer, 1778) is a hematophagous mite species, which is very commonly found in layer facilities in Europe. The economic and animal health impact of this parasite is quite important. In laying hen houses, orga- nophosphates are almost the only legally usable chemicals. Detecting a target resistance can be useful in order to limit the emergence of resistant populations. The acetylcholin- esterase (AChE) activity and the enzyme sensitivity to paraoxon was investigated in 39 field samples and compared to a susceptible reference strain (SSK). Insensitivity factor values (expressed as IC50 ratio) obtained from field isolates compared to SSK revealed some polymorphism but not exceeding a 6-fold difference. The kinetic characteristics of AChE from some field samples showed some difference in KM values for acetylthiocholine and inhibition kinetics performed with diethyl paraoxon exhibited a 5.5-fold difference in the bimolecular rate constant in one field isolate. Taken together, these data suggested that differences in AChE susceptibility to organophosphates may exist in D. gallinae but no resistant population was found.
Keywords Dermanyssus gallinae Inhibition of acetylcholinesterase Paraoxon Field isolates Organophosphate resistance
Introduction
Dermanyssus gallinae (De Geer, 1778) or the poultry red mite is a cosmopolitan hema- tophagous mite, parasitic on birds. Five life stages are known for this species (egg, larva, protonymph, deutonymph, adult), two of which need a blood meal for performing
L. Roy (&) C. Chauve T. Buronfosse Ecole Nationale Ve´te´rinaire de Lyon, Laboratoire de parasitologie, Universite´ de Lyon, 69280 Marcy-L’Etoile, France e-mail: [email protected]
J. Delaporte G. Inizan Bayer Healthcare Animal Health, Bayer Pharma S.A.S, 10A, rue Joseph Le Brix, BP 40011, 56891 Saint-Ave Cedex, France 123 20 Exp Appl Acarol (2009) 48:19–30 metamorphosis (protonymph, deutonymph; Wood 1917). Adult females need blood meals for egg maturation. The economic impact of this parasite is quite important with more or less serious direct damages, such as anemia (Kirkwood 1967; Kec¸eci et al. 2004), possible death from exsanguination, decreased egg production, but also possible transmission of certain bacterial or viral diseases (avian spirocheatosis, fowl cholera, salmonellosis, etc.) (Valiente Moro et al. 2005, 2007). Moreover, some well visible blood spots on egg shells induce a heavy financial loss with downgraded eggs. It is especially injurious in layer houses in Europe and, today, controlling the spread of these mites is an economic chal- lenge. Because of the Maximum Residue Limits in eggs, only few products are allowed for the control of D. gallinae in Europe. Insects and other arthropods have developed different mechanisms to escape to a selective pressure imposed by the use of the same insecticide. One of the adaptive mechanisms, which confer resistance to organophosphates (OPs) and carbamate pesticides, is allowed by a modification of the acetylcholinesterase (AChE), the insecticide target protein (Fournier and Mutero 1994). AChE catalyses the hydrolysis of the neurotrans- mitter, acetylcholine, thereby ending transmission of nerve impulses at the synapses of cholinergic neurones. The inhibition of this enzyme leads to paralysis and death of arthropods. Conversely, AChEs that are not inhibited by OPs and carbamates confer resistance to these pesticides. In resistant arthropods, a structural change of AChE pre- serves it from being inhibited by the OPs. This structure modification has been shown in more than 33 insect and acari species (Fournier and Mutero 1994). For instance: some crop pests such as spider mites (Stumpf et al. 2001; Tsagkarakou et al. 2002) or codling moth (Reuveny and Cohen 2004), some stored-food pests such as Psocidae (Wang et al. 2004), some potential diseases vectors such as mosquitoes (Weill et al. 2003), some ticks (Stone et al. 1976; Baxter et al. 1999; Pruett 2002), etc. The molecular basis of this resistance has been characterized for some insects (Zhu et al. 1996; Newcomb et al. 1997; Nabeshima et al. 2004) and ticks (Xu et al. 2003) and is associated with specific mutations in the ace genes. Single or multiple amino acid substitutions confer distinct catalytic properties to the mutated protein leading to a decreased sensitivity of AChE to inhibition by OPs insecticides. This phenomenon has been developed following extensive and prolonged use of these insecticide compounds. Because the frequency of D. gallinae infestations is currently increasing and that large populations can be established rapidly under favorable conditions, farmers, worried by economic losses, use chemical acaricide treatments at least in the empty chicken houses (Chauve 1998). All these factors are therefore prerequisite to expect that a resistant strain may be favored if a benefit mutation point in the D. gallinae AChE gene arised. High levels of resistance in D. gallinae have been reported for DDT and permethrin resistant mites was suspected to be involved as the main reason for the failure to control some D. gallinae populations (Zeman and Zelezny 1985; Nordenfors et al. 2001). In French laying hen houses, almost only OP compounds can be legally used, and, until recently, only between flocks. Previously, no ectoparasiticide was allowed to be used during flocks except some products composed of vegetal extracts, inert substances and some detergents, with mechanical actions. The lack of efficacy of these compounds compared to cholinesterase inhibitors incitated egg farmers to use these chemicals as soon as the poultry red mites represented a major problem in aviary systems. As a result, many populations of European mites might have been repeatedly exposed to OPs. The fast development potential of D. gallinae in layer houses conditions and the applications of OPs, legally done between flocks and sometimes illegally during flocks, to maintain 123 Exp Appl Acarol (2009) 48:19–30 21 the mite populations below economic thresholds are factors that may have facilitate the emergence of insecticide resistance in this species. Since 2007, an OP ectoparasiticide, phoxim, which can assure a 0-day withholding period for eggs, was approved by EMEA to be used, in Europe, to treat a D. gallinae infestation in poultry houses stocked with egg-laying hens (Keı¨ta et al. 2006). Thus, one can suppose that as a new OP-based product is going to be used during flocks, poultry red mites will be exposed constantly, forward favoring the emergence of resistant populations. The aim of the present study, which was conducted before the commercial authorization for using phoxim, was to use a biochemical assays for monitoring AChE in mites coming from different layer houses from different French counties in order to investigate the possible existence of resistant strains.
Materials and methods
Mites
Fourty different populations of mites were used in this study. A putative susceptible reference strain called Standard Strain Kilpinen (SSK) was kindly provided by Dr Ole Kilpinen (Lyngby, Denmark). This strain has been cultured in laboratory conditions since 1997 and has not been exposed to OPs since at least that date. The 39 other mite populations were collected in 39 independent layer houses from various counties in France. All isolates were maintained alive separately in the laboratory for few days (less than 6 days) allowing the emergence of a sufficient number of proto- nymphs to perform the biochemical assay. Thus, as mites were directly coming from farms, most of protonymphs were engorged. In order to get enough living and unengorged pro- tonymphs, females were allowed laying their eggs and eggs hatching by placing each strain at room temperature in an open box. Each box was placed into a large bowl filled with water (with a drop of a tension-active agent) so that mites cannot escape and that inter- isolate contamination is avoided. To standardize the biochemical assay as a potential diagnostic test and to avoid expected interferences due to blood meal in adults or in deutonymphs, only unengorged protonymphs were ground and AChE extracted. Typically, 200 protonymphs were placed into a 1.5 ml microcentrifuge tube and killed by freezing (-20 C for 2 h). Samples were ground in ice in the same tube containing 1.4 ml of a 10 mM pH 7.5 Tris–HCl buffer, 5.84% (m/v) NaCl, 0.4% Triton X-100 and 25 mM EDTA using a Potter’s device. AChE was extracted for 20 min at 4 C and extracted AChE was harvested in the supernatant following centrifugation at 14.000g for 10 min at 4 C. These extracts were immediately used for measuring AChE activity.
AChE assays
Basic principle
AChE activity of protonymph extracts was measured with a modification of the Ellman assay based on the enzymatic hydrolysis of acetylthiocholine iodide, ASCh, (Ellman et al. 1961). Reactions were conducted in 96-well microplates (Maxisorp, Nunc, France). Typical AChE activity of protonymph extract was assayed on 100 ll of extract mixed with 100 ll of buffering solution containing Tris-HCl 0.5 mM pH 8.0, 1.6 mM of 5-50-dithio-bis 123 22 Exp Appl Acarol (2009) 48:19–30
(2-nitro-benzoic acid) (DTNB, Sigma chemicals) and ASCh 1 mM. The change of absorbance at 410 nm was measured every 4 min for 40 min at 25 C on a Dynex mi- croplates reader. Each assay was performed in duplicate. The spontaneous ASCh hydrolysis was corrected to the signal obtained from each well by subtracting the change of absorbance occurring in wells where protonymph extracts were omitted.
Preliminary assays for the validation of the test
Because assays on the susceptibility of AChE have never been done on D. gallinae, some parameters were first checked for the validation of tests. The linearity of the reaction with time and with the enzyme amount has been explored by measuring the enzymatic activity on extracts containing various numbers of protonymphs (from 50 to 400 protonymphs) in the same amount of lysis buffer. The stability of protonymph AChE to freezing was evaluated over a 3 months period using several aliquots of protonymphs conserved at -20 C. These aliquots were regularly extracted and their AChE activity measured between day 1 and 90. Moreover, AChE activity was measured exclusively on this stage because other stages (deutonymphs or adults) may content some residual esterase activities coming from the host’s blood. However, few assays were conducted on AChE extracted from adults after 2 weeks of starvation. Measured activities were roughly in the same order of magnitude than those observed with protonymph extracts but interassay variations appeared to be important (results not shown), suggesting a possible interaction of remaining blood enzyme activities as soon as mites have had a blood meal. In order to verify that the change in the absorbance at 410 nm of protonymph extracts was dependent of AChE activity, inhibition studies were assayed. The thermal inactivation of the enzymatic activity was evaluated by measuring the residual activity of the extract after an incubation for 10 min at 50 C of the protonymph extracts. The ability of the carbamate eserine sulphate to inhibit the change in the absorbance was also evaluated by measuring the activity in the presence of 0.5 mM of eserine.
Exploration of kinetic parameters of extracted AChE
The kinetic parameters KM and Vmax of AChE extracts from the SSK strain and four isolates from field were determined with 12 different concentrations of ASCh ranging from 10 lM to 5 mM of final concentration. The AChE activity was converted to picomoles of ASCh hydrolyzed per min and per protonymph using 1.36 104 M-1 cm-1 as molar extinction coefficient. Kinetic constants KM and Vmax were obtained by linear regression after fitting a Lineweaver-Burk double-reciprocal plot of the Michaelis-Menten function.
OP inhibitory assay
To investigate whether the AChE activity, extracted from protonymphs sampled in dif- ferent laying farms, was susceptible to OP inhibition, inhibition kinetics were analyzed with the OP inhibitor diethyl paraoxon which is the form of the insecticide that irreversibly inhibits AChE. The organophosphate diethyl paraoxon [O,O-diethyl-O-(4-nitrophenyl) phosphate] was purchased from Riedel-de Haen, Seelze, Germany and was used in inhibitory assays as an AChE specific inhibitor. 123 Exp Appl Acarol (2009) 48:19–30 23
Progressive inhibition of AChE activity by OP over time
A progressive inhibition of AChE activity over time was performed in the AChE extract from the SSK strain and three additional field samples (number 6002, 6005 and 22S84). The inhibitory action of diethyl paraoxon on AChE extracts from these samples was also analyzed by calculating the bimolecular rate constant, ki, using the Aldridge method (1950). Briefly, residual AChE activity was measured as stated above for standard AChE assay except that supernatant extract was incubated with different concentrations of diethyl paraoxon (2.0 9 10-7 M, 1.0 9 10-7 M, 5.0 9 10-8 M) for various times (0, 6, 12, 19 and 30 min) prior the addition of the substrate reagent. In each assay, blank controls were done with wells without protonymph extract, in order to subtract non enzymatic hydrolysis of ASCh. The logarithm of the residual activity was plotted against the preincubation time and the bimolecular rate constant, ki, was extracted by linear regression by dividing the slope by the inhibitor concentration in accordance with Aldridge (1950).
Screening of 39 field populations
In order to explore potential variations in the susceptibility to OP inhibition of AChE extracted from protonymphs sampled in different laying farms, inhibition of AChE for each 39 field samples was measured in the presence of 1 mM ASCh by three concentra- tions of inhibitor and without any preincubation time. Paraoxon concentrations used were 8.0 9 10-7 M, 2.0 9 10-7 and 5.0 9 10-8. These concentrations were selected to inhibit between 20 and 80% of the residual AChE activity in SSK strain without any delay between the addition of diethyl paraoxon and the addition of the substrate reagent. Each sample was analyzed in two separate assays. In each duplicate, a reaction without the inhibitor was included as a control. An OP inhibitory assay in duplicate of the SSK strain was also systematically included in each plate. All the assay conditions were the same as that used for kinetics assays. The inhibitory data were analyzed by plotting residual activities with the inhibitor concentration. Inhibitory concentration 50% (IC50) value of each field sample was evaluated by linear regression and the ratio of this IC50 value with that obtained with the SSK strain in the same plate was calculated.
Statistical analysis
Kinetic parameters were compared using a non-parametric Kruskal–Wallis test.
Results
Validation of AChE extraction
The AChE activity was measured from an extract of protonymphs from the SSK strain in the supernatant and in the pellet after suspending the disrupted protonymph fragments in 100 ll of the lysis buffer. Variations in the change of absorbance obtained when incu- bations were performed with the pellet were not significantly higher than the spontaneous hydrolysis of ASCh, suggesting that no AChE activity remained present in the non- extracted fraction (results not shown). The AChE activity was linear with incubation time up to 40 min and with the amount of protonymphs extracted up to 200 protonymphs per 1.4 ml of lysis buffer (Fig. 1a). Above this quantity, the activity plateaued probably due to 123 24 Exp Appl Acarol (2009) 48:19–30
Fig. 1 Kinetic properties of A acetylcholinesterase extracted from N1 stage from Dermanyssus gallinae. a Acetylcholinesterase activity of D. gallinae extracted from different quantities of protonymphs. Each data point represents the velocity of thiocholine production from
acetylthiocholine calculated as OD value (410 nm) the OD (410 nm) changes after 5, 10 or 15 min incubation time in the presence of the substrate (1 mM). b Hydrolysis of acetylthiocholine as a fonction of incubation time when AChE Number of Protonymphs extracted from 200 protonymphs is used. c Acetylcholinesterase B activity in the presence or absence of eserine (0.5 mM) or when extracts containing acetylcholinesterase have been incubated 10 min at 50 C prior the introduction of the substrate. OD (410 nm) was measured after 40 min of incubation. Bars represent mean ± SD OD value (410 nm)
Time (min)
C OD value (410 nm)
competitive reactions. For this reason, all assays were further performed using 200 pro- tonymphs per extraction. The OD curve was linear with incubation time at least up to 40 min (Fig. 1b). The catalytic activity mediated by the protonymph extract was totally inhibited by thermal pretreatment at 50 C for 10 min or by incubation with 0.5 mM eserine sulphate, a carbamate compound that specifically inhibits AChE (Fig. 1c). 123 Exp Appl Acarol (2009) 48:19–30 25
Further, AChE stability to freezing was checked over a 90-days period. No significant loss of AChE activity was observed in protonymphs that had been stored at -20 C for up to 3 months before extraction (data not shown). Thus, this microtiter plate assay using D. gallinae protonymphs was proved to be suitable for measuring AChE activity and per- forming inhibition studies.
Kinetic parameters of AChE in Dermanyssus gallinae
The catalytic properties of D. gallinae AChE extracts have been characterized in five different strains coming from a laboratory source (SSK) or from field samples. Kinetics of the AChE extracts using ASCh as artificial substrate followed Michaelis–Menten kinetics. Kinetic parameters of the AChE extracts are shown in Table 1. Despite apparent differ- ences in KM values between isolates under test, there was no significant difference. Nevertheless, a lower Vmax value was obtained with AChE extracts from SSK and 6001 (P = 0.004) compared to the other field samples. The AChE inhibition kinetics obtained with the reference strain SSK and three field samples (numbers 6002-6005-22S84) were further characterized for their sensitivity to diethyl paraoxon. The progressive inhibition of AChE curves followed a pseudo first order kinetics (Fig. 3). The apparent bimolecular rate constants (ki) for AChE inhibition were extracted for each of these samples and values are given in Table 2. The ki values for the 6005 and 22S84 field samples were in the same range of magnitude than that obtained with AChE extracted from SSK strain whereas the bimolecular rate constant for 6002 strain was significantly different (P \ 0.01) from the sensitive reference strain (Table 2). The lower ki
Table 1 KM and Vmax values of AChE extracted from Dermanyssus gallinae protonymphs coming from SSK strain and from different field samples
-1 -1 Strain/field isolates Vmax (pmol min protonymph) KM (lM)
SSK (n = 8) 62.1 ± 8.7 36.4 ± 9.5 22S84 (n = 6) 106.5 ± 9.6 54.8 ± 18.5 35S72 (n = 3) 87.0 ± 4.1 44.7 ± 12.0 6001 (n = 3) 57.2 ± 1.1 38.7 ± 9.0 Berthet (n = 3) 105.0 ± 9.9 52.7 ± 7.6 Values are expressed as mean ± SD, numbers in parentheses represent number of independent determinations
Table 2 Bimolecular rate constants (ki) for AChE extracts inhibition by paraoxon in SSK strain and 3 field samples Strain/field isolates Bimolecular rate constant (M-1 min-1) Insensitivity factor
SSK (n = 8) 2.5 9 105 ± 1.3 9 105 22S84 (n = 6) 9.8 9 104 ± 1.3 9 104 2.5 6005 (n = 6) 1.7 9 105 ± 0.5 9 105 1.4 6002 (n = 12) 4.4 9 104 ± 2.2 9 104 5.5
Insensitivity factor is expressed as the ratio of bimolecular rate constants ki SSK/ki field sample ki values are expressed as the mean of several independent assays using at least two different diethyl paraoxon concentrations each. Number of independent assays is represented as (n=) 123 26 Exp Appl Acarol (2009) 48:19–30 value observed for the 6002 sample led to a 5.5-fold difference in the AChE insensitivity of this isolate toward paraoxon compared to AChE sensitivity in SSK strain. This was associated with two to sixfold increase of the time required to obtain 50% of AChE inhibition compared to SSK strain depending on the inhibitor concentration.
AChE inhibition screening
-7 IC50 values obtained from the SSK strain were 1.52 9 10 ± 0.18 M. IC50 values obtained from field samples ranged between 1.29 9 10-7 and 1.47 9 10-6 M. The insensitivity factor, calculated as a ratio between IC50 values obtained from field samples with that with the SSK strain, both obtained during a single assay, are represented in Fig. 2. The 39 field samples harbored a range in the IC50 ratio compared to that of SSK strain between 1 and 6. Over the 39 analyzed samples, none provided a IC50 ratio over 10.
Discussion
Validation of the tests
The results reported in the present study showed that the selected method is convenient and sensitive enough to detect subtle changes in AChE activities between field populations. AChE activity was linear up to 200 protonymphs per replicate and 40 min of incubation (Fig. 1a, b). Hydrolysis of ASCh appears to be catalytically mediated by AChE because of a strong decrease in activity when protonymphs extracts were incubated with 500 lM eserine or after thermal inactivation (Fig. 1c).
Kinetics
The kinetic properties of D. gallinae AChE in the reference strain SSK for ASCh were characterized and michaelian parameters were compared to those obtained from field samples. KM values were comparable to that obtained from mites such as Tetranychus urticae Koch (Tsagkarakou et al. 2002). It has been reported that incubations performed with Triton X-100 may significantly affect the kinetic constant KM exhibiting a competitive
6 Fig. 2 Acetylcholinesterase 29S44 inhibition studies of 39 field- 5.5 collected samples. IC50 for 6002 acetylcholinesterase inhibition by 5 diethyl paraoxon were SSK 49AR70
50 4.5 determined according to 26A52 ‘‘Material and methods’’. Each 85A68 data point represents the ratio of 4 IC50 value of the corresponding 3.5 29S34 29S62 5017 50AL71 Berthet field-collected sample divided by 22L7 56AS14
sample / IC 56S10 85S67 IC50 value of the sensitive 3 reference strain (SSK) obtained 50 2.5 22A18 35S72 in the same assay 29S61 29S52 22S59 78S79 29S43 2 6004 Ratio IC 29S41 1.5 22A37 22S84 6003 35AR85 26A50 62S33 85AL76 22S45 6001 6005 1 22S16 22S20 56S8 Gardener SSK 123 Exp Appl Acarol (2009) 48:19–30 27 inhibition (Chen et al. 2001; Rosenfeld et al. 2001). Nevertheless, we were unable to investigate the kinetic properties of the native enzyme because of a lack of efficiency in solubilizing AChE enzyme when the detergent was omitted from the lysis buffer. All tested field samples exhibited similar KM values and different Vmax values (some field isolates with increased Vmax values). It has been reported that insensitive AChE in Boophilus microplus exhibited a reduced KM value relative to the susceptible enzyme associated with a corresponding lower Vmax (Nolan and Schnitzerling 1975). Smissaert found similar results in T. urticae sensitive and resistant strains (Smissaert 1964). Conversely in the latter species, lower KM values in AChE sensitive strain associated with similar Vmax values as compared to resistant strains were reported (Stumpf et al. 2001). These modifications in kinetics parameters were considered to be the consequence of structural changes in the enzyme and are the biochemical support of the severe fitness cost that has been observed in most pop- ulations with insensitive AChE that expressed a reduction of AChE activity in synapses (Lenormand et al. 1999). In our study, the reference SSK strain and the field sample (6001) exhibited a significant lower Vmax value compared to the other samples. Although low AChE activity is a characteristic of a resistant phenotype (Lee and Bantham 1966), none of AChE inhibition studies conducted with extracts from these two populations exhibited specific characteristics of a resistant strain (Fig. 2; Table 2). Interestingly, comparisons of the kinetic parameter Vmax between the different populations clearly showed that polymorphism in AChE expression exists among isolates of D. gallinae under test. Whether it can be supposed that a strain which has been maintained in laboratory culture for several years without any acaricide treatment may led to produce an homogenous population with low intrinsic AChE activity, it is surprising to detect a field-collected population harboring the same low specific activity without any apparent disadvantage. As AChE inhibition by diethyl paraoxon follows a first-order kinetic, the bimolecular rate constant was determined for SSK and three field isolates as this parameter appears to be a much better index than the usual—but less time and mites consuming—IC50 (Aldridge and Davison 1952). The ki values obtained in D. gallinae were comparable to those described in T. urticae (Tsagkarakou et al. 2002), but differences between D. gallinae isolates remained small compared to those observed between sensitive and experimentally selected OP resistant strain of T. urticae (ki differences of 39-fold in Tsagkarakou et al. 2002). Indeed, kinetic analysis of the interaction of AChE from these three different field isolates (22S84-6005-6002) with diethyl paraoxon revealed a maximum difference between the ki values of *5–6 in the favor of the 6002 isolate (Table 2; Fig. 3) and was associated with a higher IC50 value (Fig. 2). The kinetics of AChE clearly showed that the enzyme, in this isolate, was different than those measured in other D. gallinae isolates. Nevertheless, whether both the lower ki value and the higher IC50 value evoked target site insensitivity, it is worth to note that the insensitivity factor, obtained in the 6002 isolate, remained low compared to factors observed in experimentally confirmed AChE-resistant strains in other related mites (Tsagkarakou et al. 2002). The decrease in ki value in this particular isolate should probably be interpreted as a probable decrease in AChE affinity to diethyl paraoxon even if a modification of the phosphorylation rate should not be excluded as it was shown in OP resistant B. microplus (Pruett 2002). In this species, the bimolecular rate constant was most affected by a slower rate of enzyme phosphorylation. At last, kinetic studies of AChE in field samples of D. gallinae exhibited a moderate heterogeneity in these activities that may be associated with different sensitivities to OP. The biological significance of this polymorphism and its potential impact on the control of the Poultry Red Mite in farms remains to be evaluated. In order to get an overview of this polymorphism prevalence, 39 populations from field isolates were screened and their IC50 measured. 123 28 Exp Appl Acarol (2009) 48:19–30
Fig. 3 Graph representative of the evolution of residual AChE activity from isolate 6002 and strain SSK with different incubation times and two different diethyl paraoxon concentrations. Isolate 6002 and strain SSK have been incubated with two different sets of concentrations of diethyl paraoxon (SSK: 2.0 9 10-7 and 1.0 9 10-7 M-1; 6002: 4.0 9 10-7 and 2.0 9 10-7 M-1). As inhibition is stopped by the introduction of substrate ASCh, the difference of incubation time generates different levels of inhibition with a single OP concentration. The apparent bimolecular rate ki was extracted from the regression of each slope
Screening of 39 field populations
Around 10% of populations under test exhibited a maximum 6-fold difference in the inhibitory effect of diethyl paraoxon compared to SSK strain (Fig. 2) whereas a 100-fold difference in the inhibitory effect of diethyl paraoxon is classically observed in tetranychid pest species (Stumpf et al. 2001) even if lower insensitivity factors have sometimes been reported for other OP compounds such as dichlorvos (ratios ranged between 25 and 38, Zahavi and Tahori 1970). Insensitivity factors from these 39 isolates were not as high as compared to other acari proved to be resistant. Indeed, even if AChE OP-insensitivity is defined by a slower rate of AChE inhibition in the resistant phenotype and that the insensitivity factor can vary depending on the species under test or the inhibitor compound used, insensitivity factors (IC50 ratios between a population under test and SSK from a single assay) obtained in our results appear too low to consider any field-collected pop- ulation as resistant. However, results of this study clearly showed that AChE from field-collected popula- tions of D. gallinae exhibited different susceptibilities to diethyl paraoxon. But no target resistance has been detected in isolates under test.
Conclusion
This study provides the basis for the development of diagnostic tools that can be used for management of possible AChE resistance in D. gallinae against OPs and carbamates insecticides. A screening on 39 field samples has revealed moderate differences in AChE sensitivity to paraoxon between field populations but AChE insensitivity to OPs has been 123 Exp Appl Acarol (2009) 48:19–30 29 considered to be too weak in comparison with analyses on other Arthropoda in literature. Thus, in spite of the existence of selection pressure, no important AChE insensitivity has occurred in D. gallinae. But, our results clearly show that isolates which were sampled in independent farms revealed distinct inhibition kinetics suggesting the existence of AChE polymorphism in D. gallinae. Additionally, it would be interested to test whether the polymorphism detected in field isolates in present study is selectable under laboratory conditions under elevated OP pressure.
Acknowledgments We are grateful for having provided SSK strain to O. Kilpinen and N. Hansen (Danish Institute of Agricultural Sciences, Denmark) and for their help and advices to A. Micoud, C. Brazier and C. Mottet (Service Re´gional de la Protection des Ve´ge´taux, France). We also want to thank Mehdi Gharbi (IUT, Villeurbanne, France), Jennifer Lanneau (IUT, Villeurbanne, France) and Coralie Pulido (LGTA, Saint-Genis Laval, France) for their technical help.
References
Aldridge WN (1950) Some properties of specific cholinestarase with particular reference to the mechanism of inhibition by diethyl p-nitrophenyl triphosphate (E605) and analogues. Biochem J 46:451–460 Aldridge WN, Davison AN (1952) The inhibition of erythrocyte cholinesterase by tri-esters of phosphoric acid. Biochem J 52:62–70 Baxter GD, Green P et al (1999) Detecting resistance to organophosphates and carbamates in the cattle tick Boophilus microplus, with a propoxur-based biochemical test. Exp Appl Acarol 23:907–914. doi: 10.1023/A:1006364816302 Chauve C (1998) The poultry red mite Dermanyssus gallinae (De Geer, 1778): current situation and future prospects for control. Vet Parasitol 79:239–245. doi:10.1016/S0304-4017(98)00167-8 Chen Z, Newcomb R, Forbes E et al (2001) The acetylcholinesterase gene and organophosphorus resistance in the Australian sheep blowfly, Lucilia cuprina. Insect Biochem Mol Biol 31:805–816. doi: 10.1016/S0965-1748(00)00186-7 Ellman GL, Courtney KD et al (1961) A new, rapid colorimetric determination of acetylcholinesterase activity. Biochem Pharmacol 7:88–95. doi:10.1016/0006-2952(61)90145-9 Fournier D, Mutero A (1994) Modification of the acetylcholinesterase as a mechanism of resistance to insecticides. Comp Biochem Physiol 108C:19–31 Kec¸eci T, Handemir E, Orhan G (2004) The effect of Dermanyssus gallinae infestation on hematological values and body weights of cocks. Turkiye Parazitol Derg 28:192–196 Keı¨ta A, Pagot E, Pommier P et al (2006) Efficacy of phoxim 50% E.C. (ByeMite) for treatment of Dermanyssus gallinae in laying hens under field conditions. Rev Med Vet 157:590–594 Kirkwood AC (1967) Anaemia in poultry infested with the red mite Dermanyssus gallinae. Vet Rec 80(17):514–516 Lee RM, Bantham P (1966) The activity and organophosphate inhibition of cholinesterase from susceptible and resistant ticks (Acari). Entomol Exp Appl 9:13–24. doi:10.1007/BF00341156 Lenormand T, Bourguet D, Guillemaud T et al (1999) Tracking the evolution of insecticide resistance in the mosquito Culex pipiens. Nature 400:861–864. doi:10.1038/23685 Nabeshima T, Mori A, Kozaki T et al (2004) An amino acid substitution attributable to insecticide- insensitivity of acetylcholinesterase in a Japanese encephalitis vector mosquito, Culex tritaeniorhyn- chus. Biochem Biophys Res Commun 313:794–801. doi:10.1016/j.bbrc.2003.11.141 Newcomb RD, Campbell PM, Russel RJ et al (1997) cDNA cloning, Baculovirus-expression and kinetic properties of the esterase, E3, involved in organophosphorus resistance in Lucilia cuprina. Insect Biochem Mol Biol 27:15–25. doi:10.1016/S0965-1748(96)00065-3 Nolan J, Schnitzerling HJ (1975) Characterisation of acetylcholinesterases of acaricide resistant and sus- ceptible strains of the cattle tick Boophilus microplus. I. Extraction of the critical component and comparison with enzyme from other sources. Pest Biochem Physiol 5:178–188 Nordenfors H, Ho¨glund J, Tauson R et al (2001) Effect of permethrin impregnated plastic strips on Dermanyssus gallinae in loose-housing systems for laying hens. Vet Parasitol 102:121–131. doi: 10.1016/S0304-4017(01)00528-3 Pruett JH (2002) Comparative inhibition kinetics for acetylcholinesterases extracted from organophosphate resistant and susceptible strains of Boophilus microplus (Acari:Ixodidae). J Econ Entomol 95:1239– 1244 123 30 Exp Appl Acarol (2009) 48:19–30
Reuveny H, Cohen E (2004) Evaluation of mechanisms of azinphos-methyl resistance in the codling moth Cydia pomonella (L.). Insect Biochem Physiol 57:92–100. doi:10.1002/arch.20016 Rosenfeld C, Kousba A, Sultatos LG (2001) Interactions of rat brain acetylcholinesterase with the detergent Triton X-100 and the organophosphate paraoxon. Toxicol Sci 63:208–213. doi:10.1093/toxsci/ 63.2.208 Smissaert HR (1964) Cholinesterase inhibition in spider mites susceptible and resistant to organophosphates. Science 143:129–134. doi:10.1126/science.143.3602.129 Stone BF, Nolan J, Schuntner CA (1976) Biochemical genetics of resistance to organophosphorus acaricides in three strains of the cattle tick, Boophilus microplus. Aust J Biol Sci 29:265–279 Stumpf N, Zebitz C, Kraus W et al (2001) Resistance to organophosphates and biochemical genotyping of acetylcholinesterases in Tetranychus urticae (Acari: Tetranychidae). Pestic Biochem Physiol 69:131– 142. doi:10.1006/pest.2000.2516 Tsagkarakou A, Pasteur N, Cuany A et al (2002) Mechanisms of resistance to organophosphates in Tetr- anychus urticae (Acari: Tetranychidae) from Greece. Insect Biochem Mol Biol 32:417–424. doi: 10.1016/S0965-1748(01)00118-7 Valiente Moro C, Chauve C, Zenner L (2005) Vectorial role of some dermanyssoid mites (Acari, Meso- stigmata, Dermanyssoidea). Parasite 12:99–109 Valiente Moro C, Fravalo P, Amelot M et al (2007) Colonization and invasion in chicks experimentally infected with Dermanyssus gallinae contaminated by Salmonella enteritidis. Avian Pathol 36:307–311. doi:10.1080/03079450701460484 Wang JJ, Cheng WX et al (2004) The effect of the insecticide dichlorvos on esterase extracted from the psocids, Liposcelis bostrychophila and L. entomophila. J Insect Sci 4:23–27 Weill M, Lutfalla G, Mogensen K et al (2003) Comparative genomics: Insecticide resistance in mosquito vectors. Nature 423:136–137. doi:10.1038/423136b Wood HP (1917) The chicken mite: its life history and habits United States Department of Agriculture, Washington, DC. Bull 553:1–14 Xu G, Fang QQ, Keirans JE et al (2003) Cloning and sequencing of putative acetylcholinesterase cDNAs from the American dog tick, Dermacentor variabilis, and the brown dog tick, Rhipicephalus san- guineus (Acar : Ixodidae). J Med Ent 40:890–896 Zahavi M, Tahori AS (1970) Sensitivity of acetylcholinesterase in spider mites to organophosphorous compounds. Biochem Pharmacol 19:219–225. doi:10.1016/0006-2952(70)90342-4 Zeman P, Zˇelezny` J (1985) The susceptibility of the poultry red mite, Dermanyssus gallinae (De Geer, 1778), to some acaricides under laboratory conditions. Exp Appl Acarol 1:17–22. doi:10.1007/ F01262196 Zhu KY, Lee SH, Clarck JM (1996) A point mutation of acetylcholinesterase associated with azinphosm- ethyl resistance and reduced fitness in colorado potato beetle. Pestic Biochem Physiol 55:100–108. doi: 10.1006/pest.1996.0039
123 Exp Appl Acarol (2009) 48:63–80 DOI 10.1007/s10493-009-9239-1
Candidate predators for biological control of the poultry red mite Dermanyssus gallinae
Izabela Lesna · Peter Wolfs · Farid Faraji · Lise Roy · Jan Komdeur · Maurice W. Sabelis
Received: 10 October 2008 / Accepted: 8 January 2009 / Published online: 30 January 2009 © Springer Science+Business Media B.V. 2009
Abstract The poultry red mite, Dermanyssus gallinae, is currently a signiWcant pest in the poultry industry in Europe. Biological control by the introduction of predatory mites is one of the various options for controlling poultry red mites. Here, we present the Wrst results of an attempt to identify potential predators by surveying the mite fauna of Euro- pean starling (Sturnus vulgaris) nests, by assessing their ability to feed on poultry red mites and by testing for their inability to extract blood from bird hosts, i.e., newly hatched, young starlings and chickens. Two genuine predators of poultry red mites are identiWed: Hypoa- spis aculeifer and Androlaelaps casalis. A review of the literature shows that some authors suspected the latter species to parasitize on the blood of birds and mammals, but they did not provide experimental evidence for these feeding habits and/or overlooked published evidence showing the reverse. We advocate careful analysis of the trophic structure of arthropods inhabiting bird nests as a basis for identifying candidate predators for control of poultry red mites.
Keywords Biological control · Ectoparasite · Poultry red mite · Dermanyssus gallinae · Predatory mites · Androlaelaps casalis · Hypoaspis aculeifer · European starling · Sturnus vulgaris · Chicken · Gallus gallus · Trophic structure · Bird nest · Poultry house
I. Lesna · M. W. Sabelis Section Population Biology, Institute for Biodiversity and Ecosystem Dynamics, University of Amsterdam, Kruislaan 320, 1098 SM Amsterdam, The Netherlands
I. Lesna (&) · P. Wolfs · J. Komdeur Animal Ecology Group, Centre for Ecology and Evolution, University of Groningen, P.O. Box 14, 9750 AA Haren, The Netherlands e-mail: [email protected]; [email protected]
F. Faraji MITOX Trial Management BV, P.O. box 92260, 1090 AG Amsterdam, The Netherlands
L. Roy Laboratoire de Parasitologie et Maladies Parasitaires, Ecole Nationale Vétérinaire de Lyon, Marcy l’Etoile, France 1 C 64 Exp Appl Acarol (2009) 48:63–80
Introduction
The poultry red mite, Dermanyssus gallinae (De Geer) (Acari: Dermanyssidae), is a blood- sucking ectoparasite in nests of birds and small mammals. It is of economic importance as a worldwide pest in the poultry industry (Axtell and Arends 1990). This holds especially for ‘laying hen’ houses because the period of egg laying and breeding of domesticated chick- ens (Gallus gallus) is long (Maurer et al. 1993; Emous et al. 2005) relative to the time required for poultry red mites to double their population size (5.9 days at 25°C; Maurer and Baumgärtner 1992, 1994). In the Netherlands, outbreaks occurred infrequently in the past (mainly during summer), but now these occur throughout the year and in virtually all ‘laying hen’ (but also ‘broiler breeder’, ‘rearing hen’ and ‘parent stock’) houses. This increased incidence may be due to modern farming systems (e.g., constant climate) and perhaps also due to pesticide resistance arising from intensive chemical control. In addi- tion, more strict allowance regulations have limited the number of pesticides available for control. As eggs are produced for human consumption, pesticides have to meet strict stan- dards for food safety: (1) no residuals and (2) rapid breakdown into non-harmful compo- nents. Selection of eVective chemicals is further complicated by the demand that pesticides should not harm bird health or bird welfare. Exposure of chickens to chemical sprays is hard to avoid, even though poultry red mites do not stay on their host permanently. This is because poultry red mites spend most of their time in the vicinity of the chicken and then they hide in crevices and other narrow places in the farm structure, where they are hard to target by pesticide sprays. The currently admitted pesticides do not suYce to eradicate poultry red mites. Damage due to poultry red mites involves anaemia, increase in death risks and food demands, reduced time available for resting, decrease in disease resistance and egg-laying, and reduced egg quality (downgraded eggs due to egg shell with blood spots arising from squashed poultry red mites) (Emous et al. 2005). Based on a recent inquiry among farmers, the total costs from damage and control measures to the whole egg industry in The Nether- lands are estimated to be 11 million Euros per year (Emous et al. 2005). This underesti- mates real costs because poultry red mites may vector disease agents of poultry (Valiente Moro et al. 2005, 2007) and they are a source of allergens causing dermatological problems to farmers and veterinarians (Rosen et al. 2002; Beck 1999). Because current control methods are not suYciently eVective (Emous et al. 2005), we aimed to develop new methods of poultry red mite control by the use of their natural ene- mies, in particular predatory mites. This approach was pioneered by BuVoni et al. (1995, 1997) and Maurer and Hertzberg (2001). They reported the spontaneous occurrence of the predatory mite Cheyletus eruditus (Schrank) in the litter of poultry houses in Switzerland (for similar Wndings in Egypt, Mexico and UK, see: Abo-Taka 1996; Quintero and Acevedo 1984; Brady 1970a, b) and observed this mite feeding on juvenile poultry red mites. Releases of this predatory mite in poultry houses turned out not to yield control of poultry red mites, however. We pursued another approach to develop biological control methods by identifying predators of poultry red mites in their natural habitat. We analysed the food web structure of bird nests that have natural infestations of D. gallinae. Inventories of arthropods inhabiting bird nests have revealed a community structure involving bird parasites, microbi- vores and predators that may feed on them (e.g., Philips and Dindal 1979; Philips et al. 1989; Putatunda et al. 1989; Gupta and Paul 1989; Burtt et al. 1991; Lundqvist 1995; Kris- tofík et al. 1996; Philips 2000). Feeding at more than one trophic level (omnivory) may also be possible, but has not yet been shown for mites inhabiting bird nests. Some of the species are found exclusively in nests (nidicolous species), others are opportunistic visitors (e.g., 1 C Exp Appl Acarol (2009) 48:63–80 65 edaphic species) and yet others (plant-dwelling species) end up in nests via plant material birds use for nest construction, and perhaps also via the green material birds use for court- ship, for chemical control of ectoparasites or as a drug to boost their immune response (Clark 1991; Brouwer and Komdeur 2004; Gwinner and Berger 2005; Veiga et al. 2006). For the arthropods inhabiting a nest to establish a system with more trophic levels will require time. Hence, elaborate micro-ecosystems of arthropods are more likely to occur in association with birds that re-use nesting sites, such as European starlings (Sturnus vulgaris Linnaeus 1758). Our approach was therefore (1) to identify the mites inhabiting poultry houses and nests of European starlings, (2) to identify potential predators based on the liter- ature and based on their numerical associations with D. gallinae, (3) to assess their preda- tory activity with respect to the blood parasite D. gallinae and microbivores, such as astigmatic mites, and Wnally (4) to test whether these candidate predators can switch to blood feeding on starlings and chicken in absence of prey (in which case they are omnivores because they feed at more than one trophic level).
Materials and methods
Mite inventory of poultry houses
Surveys of the mite fauna inhabiting four poultry farms in the provinces Brabant and Limburg (The Netherlands) were made every 2 weeks in the period from October 2007 to July 2008. Per sampling date and poultry house 20 samples were taken by Nordenfors traps, i.e., traps made of 3 mm thick, 140 £ 100 cm, corrugated cardboard (Nordenfors and Chirico 2001). These traps were placed in laying nests and on perches, poles and walls. In addition, a variable number of samples from poultry house litter and from conspicuous D. gallinae aggregations were taken. Samples were inspected Wrst under a binocular microscope and any mite suspected to be diVer- ent from D. gallinae was collected in vials with alcohol, mounted in modiWed Hoyer’s medium (Faraji and Bakker 2008) on a microscope slide and then identiWed. If available, at least 10 voucher specimens per species were maintained for later inspection. Estimates of abundance were classiWed as rare (1–5 individuals/sample), common and never abundant (5–100/sample), common and sometimes abundant (>100/sample).
Mite inventory of starling nests
A total of 106 nest boxes, at least 6 m apart and 2.5 m above ground, were at our disposal at Vosbergen estate (Eelde-Paterswolde, The Netherlands). Because starlings re-use nest boxes and the mite fauna may become more diverse with time, the occupation of the nest boxes by breeding starlings was recorded in 2006 and the presence of old nest material was assessed in March 2007. A total of 29 nest boxes were occupied by starlings in April 2007, 14 of which had also been occupied in 2006. A large proportion of old nest material (from 2006 or earlier) was collected from each of all 29 nest boxes to assess the arthropod fauna. Nest boxes were not cleaned, however, to allow a signiWcant proportion of the mite fauna to hide in grooves, cracks and crevices in the nest box. Within a week after Xedging of the young starlings in June 2007, nest material was collected from the nest boxes. After transferring samples from the old (March 2007) and new (June 2007) nest material to Berlese funnels, the mites were collected in vials with alcohol over a period of 3–4 days, mounted in modiWed Hoyer’s medium (Faraji and Bakker 2008) on microscope slides and then identiWed. If available, at least ten voucher specimens per species were maintained for later inspection. 1 C 66 Exp Appl Acarol (2009) 48:63–80
Species-speciWc densities per nest were estimated and subsequently classiWed in the following ranges: 0–1; 2–10; 11–100; 101–1,000; 1,001–10,000; 10,001–100,000 individu- als per nest. For statistical analyses these ranges were arbitrarily transformed into single discrete numbers, representing categorical densities: 1; 5; 50; 500; 5,000; 50,000.
Statistical analysis of mite–mite associations
Categorical densities were incorporated in statistical analyses as log10 transformed (cate- gorical) variables. The statistical analysis was part of a more comprehensive multilevel modelling approach (MLwiN 2.10 beta 5 package; Rabash et al. 2008), with three hierar- chical levels involving nest-box-group, nest box and nestling. A stepwise backward elimi- nation was performed manually by removing the least signiWcant independent Wxed variable from the model. As part of this modelling exercise, relationships of ‘putative predator’ densities (individ- uals per nest) were tested using a model with the variables occupancy history of nest box (occupied in 2006 or not) and D. gallinae density (individuals per nest) as well as three variables describing the state of the starlings occupying the nest box (brood size, hatching date, female quality; for deWnitions see Brouwer and Komdeur 2004). Since the emphasis in this article is on the inter-relationships between nest-inhabiting mites, we refrain here from describing the starling-related variables and discussing the results ensuing. This will be published elsewhere (P. Wolfs et al., in prep.).
Predation tests
Based on the nest inventory and the analysis of mite–mite associations, predatory activity was assessed for adult females of Androlaelaps casalis (Berlese). For reasons of compari- son, we chose females of Hypoaspis aculeifer (Canestrini) because (1) it was occasionally found in starling nests, (2) it was observed to feed on D. gallinae (I. Lesna, pers. obs.) and (3) it was reared in our laboratory since 1991 (Lesna et al. 1995), and therefore readily available. Both A. casalis (obtained from starling nests in 2007) and H. aculeifer (obtained from lily Welds in 1991) were cultured on a diet of Tyrophagus putrescentiae (Schrank), which in turn was reared on dry yeast Xakes (22°C, 70% RH). One week before the preda- tion test the cultures of both species of predatory mites were provided with ample prey to satiate the predators. Females nearing oviposition (i.e., showing a full grown terminal oocyte in their opisthosoma) were taken from the culture and transferred to vials (3 cm diameter, 4 cm high) with a moistened bottom of plaster of Paris mixed with charcoal. The vials had ten individuals of D. gallinae (mobile juveniles and adults). After transfer, the vials were tightly closed by a lid with an opening sealed with mite-proof gauze and placed in a climate room at 22°C, 70% RH and total darkness. After 24 h, numbers of live and dead D. gallinae were counted and all individuals were replaced by fresh ones to achieve the same starting density for a second day of the predation test. The predation experiments were replicated simultaneously in ten vials for each species and each day. Control experi- ments with vials containing D. gallinae alone were also carried out. Student t-tests for com- parison of means were applied to detect diVerences in predation between the two species and between the 2 days of the experiment. Apart from the predation assessment over two consecutive days we also made direct observations of more than 100 predation events and together with Prof. Urs Wyss (Univer- sity of Kiel, Germany) we recorded part of these by the aid of a video-equipped binocular microscope. Stills of the video-records are included in this article (Figs. 3, 4). 1 C Exp Appl Acarol (2009) 48:63–80 67
Haematophagy tests
To test whether mites shown to have a capacity to prey upon D. gallinae also have a capac- ity to extract blood directly from the host of D. gallinae, we carried out two experiments one with starling and another with chicken. Three-day-old starlings and ten-day-old chick- ens were used to oVer a bird stage most vulnerable to haematophagy (due to their thin skin and hence more accessible blood vessels). Moreover, given their less dense feather cover it was easier to observe bite marks, as well as mites especially on the starling host. The experiments with young starlings were carried out in a brood incubator at 32–34°C (total darkness), installed at the Weld station in Vosbergen estate. Due to the proximity of the Weld station to the starling-occupied nest boxes, we were able to minimize the time between brood removal and their introduction into the incubator, as well as their reintroduction into their original nest. The young starlings were away from their nest no longer than 2.5 h. The experiments were carried out for 2 h during mid-day (13:00–15:00) to ensure that the young starlings had been fed by their parents before the experiment and that they would be fed by their parents before night. The young starlings were introduced each into a separate con- tainer (300 ml, 6 cm diameter) closed by a lid with a hole sealed with mite-proof gauze. Pieces of Wlter paper were provided on the bottom of the vial as a means to absorb moisture and to accommodate the young birds. Per container, 20 female mites of either A. casalis, H. aculeifer or D. gallinae were released that had been deprived of food for at least 3 days prior to the experiment. These three treatments were replicated Wve times and all replicates were carried out simultaneously. Bite marks on the young starling, gut colouration of the mites (as a bloodmeal indicator) and the presence of mites on and oV the bird were checked just before and immediately after the 2 h experiment. The experiments with young chickens were carried out from 18:00 to 12:00 next day in a climate room at 26 § 2°C (darkness from 19:00 to 10:00 next day), at the experimental Farm Laverdonk (Heeswijk, The Netherlands). The experiment was carried out mostly during the night because D. gallinae is most active at night when the young chickens are sleeping. The young chickens were brought in cages with food and water in the climate room, the day before the experiment. At the start of the experiment the chickens were intro- duced each into a separate container (3 l volume, 17 cm diameter) closed by a lid with a hole sealed with mite-proof gauze and they had no access to food and water during the time spent in the container. Pieces of soft cardboard (from egg baskets) were provided on the bottom of the container as a means to absorb moisture and to accommodate the young birds. Per con- tainer 20 mites, nymphs and females of either A. casalis, H. aculeifer or D. gallinae, were released that had been deprived of food for at least 3 days prior to the experiment. These three treatments were replicated ten times. Bite marks on the young chickens, gut colour- ation of the mites and the presence of mites on and oV the chickens were checked before and after the 18 h experiment.
Results
Mite inventory of poultry houses and starling nests
During summer time, the densities of the poultry red mite, D. gallinae, increased dramati- cally (up to 60.000 mobile stages per Nordenfors trap) with the time the laying hen Xocks stayed in the poultry house (I. Lesna, pers. obs.). In individual starling nests, densities of D. gallinae could increase from a few individuals appearing at the time young starlings 1 C 68 Exp Appl Acarol (2009) 48:63–80 hatch to more than 30.000 at the end of the starling’s breeding period (I. Lesna, pers. obs.). As argued in the Appendix, populations of D. gallinae from starling nests and poultry houses are conspeciWc. Blood-feeding mites other than D. gallinae were not found in poul- try houses and starling nests, except for a few individuals of Ornithonyssus sylviarum (Canestrini and Fanzago) in a single starling nest (Roy et al. 2009). Astigmatic mites varied in abundance largely depending on the presence of certain plant-derived materials, such as straw (collected from farmland nearby) and seeds in star- ling nests as well as chicken-feed in poultry houses. The species composition in starling nests was more diverse than in poultry houses (Table 1). Whereas three species of glycy- phagid and histiostomatid species prevailed in starling nests, it was almost exclusively, the acarid mite, Tyrophagus putrescentiae (Schrank), that was found in poultry houses. The astigmatic mites are all thought to be microbivores and may serve as the main or alternative prey for various predatory mites.
Table 1 List of Acari found in association with starling nests (estate Vosbergen, Eelde-Paterswolde, Gron- ingen) and in litter of poultry houses (Brabant, Limburg, The Netherlands)
Order Family Species Starling Nest Poultry House
Mesostigmata Ascidae Blattisocius keegani Fox – *** Proctolaelaps pygmaeus (Müller) – * Proctolaelaps sp.a *– Zerconopsis remiger (Kramer) *** – Dermanyssidae Dermanyssus gallinae (De Geer)b **** **** Digamasellidae Dendrolaelaps fallax (Leitner) * – Laelapidae Androlaelaps casalis (Berlese) *** ** Hypoaspis aculeifer (Canestrini) * – Macrochelidae Macrocheles ancyleus Krauss * – Macronyssidae Ornithonyssus sylviarum *– (Canestrini and Fanzago) Parasitidae Parasitellus fucorum De Geerc *– Polyaspididae Uroseius acuminatus (Koch) – * Prostigmata Cheyletidae Cheyletus eruditus (Schrank) ** ** Tydeidae Lorryia reticulata (Oudemans) * – Astigmata Acaridae Aleuroglyphus ovatus (Troupeau) * – Tyrophagus putrescentiae (Schrank) – ** Tyrophagus longior (Gervais) * – Sancassania sp.–* Glycyphagidae Glycyphagus domesticus (De Geer) ** – Lepidoglyphus destructor (Schrank) ** – Histiostomatidae Myianoetus sp.d ** – Pyroglyphidae Dermatophagoides evansi Fain, –* Hughes and Johnston Winterschmidtiidae Saproglyphus sp.d *–
* Rare, ** common, yet never abundant, *** common and sometimes abundant, **** very abundant. Mount- ed voucher specimens are available on request to I. Lesna or F. Faraji a Possibly new species, currently being described (F. Faraji) b ConspeciWc specimens from starling nests and poultry houses (L. Roy) c A single nymph positively identiWed d UnidentiWed species 1 C Exp Appl Acarol (2009) 48:63–80 69
Prostigmatic mites were represented almost exclusively by Cheyletus eruditus in both starling nests and poultry houses (Table 1). This species is known to be a predator of diVerent species of astigmatic mites and can probably feed on D. gallinae (Maurer and Hertzberg 2001). We also noted dark-red coloured individuals collected from poultry houses where D. gallinae was very abundant. In starling nests, C. eruditus tended to be somewhat more abundant when there were more astigmatic mites, but the available data did not allow statistical analyses to test this claim. After nestlings had Xedged in June 2007, starling nests harboured mesostigmatic mites (other than D. gallinae) that were represented most frequently by Androlaelaps casalis and Zerconopsis remiger (Kramer) (Table 1). The latter species was absent in nest boxes before the start of the breeding season (March 2007), but A. casalis was then found frequently and abundantly in old nest material in ca. 80% of the nest boxes that had been occupied by star- lings in 2006 (as opposed to ca. 30% in nest boxes that had not been occupied in the 2006). In some of the poultry houses sampled, A. casalis was also found in considerable numbers, Z. remiger was absent and Blattisocius keegani Fox was relatively the most abundant species (Table 1). Each of these four species can feed on astigmatic mites, such as Acarus siro (Linnaeus) and T. putrescentiae (I. Lesna, pers. obs.). Their potential to interfere with D. gallinae is further explored below. Strikingly, the mesostigmatic genera harbouring gen- eralist predators, such as Hypoaspis spp., Macrocheles spp. and Parasitus spp., were rarely found in starling nests as well as in poultry houses.
Mite–mite associations
Based on visual inspection of scatter diagrams of the data obtained directly after the starling nestlings had Xedged, we detected the following tendencies: densities of A. casalis and Z. rem- iger were relatively high when D. gallinae density was low, whereas densities of A. casalis and Z. remiger were low, when D. gallinae density was high. Because—just before nest build- ing (March 2007)—A. casalis was more likely to be found in nest boxes that had nest material from 2006 (whereas then Z. remiger was absent), we further explored the negative relation between A. casalis and D. gallinae densities by taking the history of nest-box occupancy into account. Post-Xedging densities of A. casalis in starling nests in June 2007 were signiWcantly higher in nest boxes that had been occupied, than in nests that had no starling nest in 2006 (Fig. 1; Wald test P =0.013). Post-Xedging densities of D. gallinae were not signiWcantly diVerent between nest boxes that had been occupied in 2006 and those that had not been occu- pied in 2006 (Fig. 1; Wald test P = 0.333). Most strikingly, postXedging densities of A. casalis and D. gallinae in June 2007 showed a signiWcant negative relationship in nest boxes that had a starling nest in 2006 (Fig. 2; Wald test P < 0.001), whereas there was no signiWcant relation- ship in nest boxes that had not been occupied in 2006 (Fig. 2; Wald test P =0.414). This negative relationship requires an interpretation. It points at some form of interfer- ence between A. casalis and D. gallinae. Hence, one would expect variation in numbers among nests to be more reduced under conditions where the two interacting mite species are more likely to co-occur from the start of the starling’s breeding period (provided that mite immigration rates are low compared to mite growth rates in the nest). These condi- tions may apply to nest boxes occupied by starlings in the previous year for the following reasons: (1) in these nests A. casalis is more likely to be present at the start of the starling’s breeding period, and (2) the growth rates of A. casalis and D. gallinae are similar and most likely high (relative to migration into the nests). We hypothesize that these conditions prompted the signiWcant negative relationship in nest boxes occupied in 2006 as well as the absence of such a relationship in nest boxes unoccupied in 2006. 1 C 70 Exp Appl Acarol (2009) 48:63–80
6 ** ** 5
4
3 transformed)
10 2 (log
Mite density per nest box 1
0
Non-occupied Occupied Occupancy history
Fig. 1 Mite densities in nest boxes with diVerent occupancy history. Grey bars represent Dermanyssus gal- linae densities and white bars represent Androlaelaps casalis densities. Data for these densities were sepa- rated according to nest occupancy history; nest boxes which were not occupied (n =15 for D. gallinae; n =14 for A. casalis) and were occupied (n = 14 for both) by starlings in the previous breeding season (2006). The boxes in the Wgure represent the interquartile range, with the line within being the median. The error bars rep- resent the 10–90th quartile range. Dots (᭹) represent data points interpreted as outliers. Note that the median (= 0.70) of D. gallinae in nest boxes occupied in the previous year coincides with the 10th quartile range. Sig- niWcant diVerences between A. casalis densities from the N · O. category and those from the O. category are indicated by ** above the data ranges (Wald test; slope is equal to 1.235 with SE = 0.498 and is signiWcantly diVerent from zero at P = 0.013). DiVerences between D. gallinae densities from the two categories were not signiWcant (Wald test; slope is equal to ¡0.519 with SE = 0.936 and is not signiWcantly diVerent from zero at P = 0.333)
The observed negative relationship between A. casalis and D. gallinae densities may arise from two distinct mechanisms: (1) competition for the same food resource (bird blood), (2) predation of one mite species on the other, or (3) a combination of competition and predation. To distinguish between these mechanisms, the most simple, Wrst approach is to test whether A. casalis can feed on D. gallinae and whether it can extract blood from its bird host. The results of these predation and haematophagy tests are given below.
Predation tests
Predation tests were carried out with two species of mesostigmatic mites, one commonly found in starling nests and poultry houses, i.e., A. casalis, and one rarely found in these environments, i.e., H. aculeifer. Out of ten replicate experiments one female of H. aculeifer could not be retrieved after the second day (hence 9 replicates remained). For A. casalis two females could not be retrieved after the Wrst day (hence 8 replicates remained). Control experiments with only D. gallinae did not show any D. gallinae mortality during the 2 days and are therefore left out of further analysis. The results of the experiments with putative predators of D. gallinae (Table 2) showed that the number of dead D. gallinae did not diVer between day 1 and 2 of the experiment for H. aculeifer (t-test; P = 0.35) and for A. casalis (t-test; P = 0.23), but revealed signiWcant diVerences between H. aculeifer and A. casalis treatments on day 1 (P = 0.0012) and day 2 (P = 0.02): the number of dead prey under exposure of H. aculeifer females was 1.8–2.1 times higher than that under exposure of A. casalis females. Oviposition was observed in several replicates, on both days for H. acule- ifer and A. casalis. We refrained from quantifying oviposition because A. casalis and 1 C Exp Appl Acarol (2009) 48:63–80 71
5
4
3 densities
transformed) 2 10 A. casalis (log 1
0
051234 D. gallinae densities (log10 transformed)
Fig. 2 Densities of Androlaelaps casalis plotted against Dermanyssus gallinae. Data for these densities were separated according to nest box occupancy history; nest boxes that were occupied (᭹)(n = 14) and were not occupied (᭺) (n = 14) in the previous breeding season (2006). The continuous line with negative slope rep- W resents the signi cant relationship between A. casalis densities (log10 transformed) and D. gallinae densities ¡ (log10 transformed) in nest boxes that were occupied in 2006 (Wald test; slope is equal to 1.016 with SE = 0.169 and is signiWcantly diVerent from zero at P < 0.001). For nests that were not occupied in 2006 the slope of the regression was not signiWcantly diVerent from zero (Wald test; slope is equal to ¡0.210 with SE = 0.257 and is not signiWcantly diVerent from zero at the 5% level since P = 0.414)
Table 2 Assessment of the rate of predation on poultry red mites (juveniles and adults) by females of two species of mesostigmatic mites during two consecutive days, following their rearing on a diet of astigmatic mites (Tyrophagus putrescentiae)
Species Number of red mites killed per day
Day 1 Day 2
Mean SD Range n Mean SD Range n
Hypoaspis aculeifer 5.5ax 1.7 3–8 10 6.4ax 2.5 3–10 9 Androlaelaps casalis 2.6ay 1.3 1–4 8 3.6ay 1.8 1–6 8
Climate room conditions were 22°C, 70% RH and total darkness; n = number of replicates; SD = Standard Deviation. SigniWcant diVerences between means of two samples according to Student t-tests are indicated by diVerent letters following the mean (between days: a, b; between species: x, y)
H. aculeifer have a strong tendency to hide eggs in small holes in the layer of plaster of Paris, and there were indications for egg retention in A. casalis females, a phenomenon known to occur in mesostigmatic mites under unfavourable conditions and reported for A. casalis by McKinley (1963). Direct observations of attacks on D. gallinae individuals and ingestion of their body Xuids were obtained for H. aculeifer (Fig. 3) and A. casalis (Fig. 4) during prey mortality assessments in the above tests and in more than 100 other predation tests. Attacks were observed on eggs and all mobile stages of D. gallinae. Large nymphs and adults were most frequently pierced at their Xanks just behind the gnathosoma. Adults of D. gallinae were not sucked dry by the predatory mites. Instead they were only partially consumed. How- ever, they were invariably leaking body Xuids through the wound, were immobilized and 1 C 72 Exp Appl Acarol (2009) 48:63–80
Fig. 3 Adult female of Hypoaspis aculeifer feeding on a nymph of the poultry red mite. Note that the gut of H. aculeifer is visible through the integument and starts to become dark-coloured (red-brown). Still from a video record made by Urs Wyss and Izabela Lesna. (Color Wgure online)
Fig. 4 Adult female of Androlaelaps casalis after feeding on a poultry red mite nymph, which causes their gut—visible through the integument—to turn red-brown (dark-coloured in this picture). Still from a video record made by Urs Wyss and Izabela Lesna. (Color Wgure online) ultimately died. Due to the transparency of their integuments, body Xuids were seen to move from the victim into the gut of H. aculeifer and A. casalis females, thereby causing the gut to assume a red-brownish colour (Figs. 3, 4). Such a change in gut colour does not occur when they fed on eggs or on unfed individuals of D. gallinae.
Haematophagy tests
Tests with starved A. casalis showed no evidence for haematophagy during 2 h of exposure to hatchlings of starlings at 32–34°C (Table 3). None of the mites were found on the host, none of them exhibited gut colouration (as in Fig. 4) and the host had no bite marks. In 1 C Exp Appl Acarol (2009) 48:63–80 73
Table 3 Replicated (n = 5) experiments to test whether starved females of two mite species feed on the blood of starling hatchlings (3 days after hatching)
Mite species Replicate number Number recovered % with coloured gut
Dermanyssus gallinae 11753 21258 31759 41968 51782 Androlaelaps casalis 1170 2160 3170 4180 5150
Shown are the number of mites recovered after 2 h from the 20 individuals initially released per container and the percentage of these recovered mites with a red-brown coloured gut. Brood incubator conditions were 32–34°C, RH >50% and no daylight contrast, tests involving D. gallinae yielded ca. 10% of the mites on host, 50–80% of the mites with red-brown coloured guts (indicating a fresh bloodmeal) and some hosts with bite marks (Table 3). Unfortunately, H. aculeifer did not survive the conditions of this experi- ment. Separate trials showed that temperatures above 30°C are detrimental to survival of H. aculeifer. For this reason the haematophagy tests with young chicken were carried out at a lower, yet bird-friendly temperature (26 § 2°C). These tests (Table 4) showed that both H. aculeifer and A. casalis females cannot obtain blood during 18 h of exposure to young chicken. In contrast, 50–90% of the D. gallinae that were recovered had red-coloured guts. Possibly, because the light had been switched on 2 h before collecting the mites from the containers, none of the D. gallinae mites were found on the hosts. Bite marks on the young chicken were noted in only few cases, but they were not easy to observe because 10-day- old chicken possess already a more dense feather cover. Not all of the 20 mites released per container were recovered at the end of the experi- ments. In the tests with starling hatchlings 12–19 D. gallinae and 15–19 A. casalis were recovered. The missing individuals are most likely present in or under the starling’s drop- pings, where part of them may have gone unnoticed. Escape during inspection and hiding on the host is quite unlikely for these experiments. In the tests with young chickens rela- tively more mites (especially D. gallinae and A. casalis) were missing for a variety of rea- sons, the most likely of which was that they were hiding in the cardbox structure and chicken faeces on the bottom of the containers. Probably due to the relatively larger size of H. aculeifer retrieval of released mites was less of a problem (10–18 mites recovered, as opposed to 4–14 for the other 2 species).
Discussion
Our approach was to explore species of putative D. gallinae predators that occur in associ- ation with D. gallinae in bird (starling) nests under natural conditions, and those that occur spontaneously in poultry farms in which D. gallinae is a pest. Two species, one laelapid (Mesostigmata) and one cheyletid (Prostigmata), were found more or less frequently in large numbers in both environments: Androlaelaps casalis and Cheyletus eruditus (Table 1). Two other species, both ascids (Mesostigmata), were found in only one of the 1 C 74 Exp Appl Acarol (2009) 48:63–80
Table 4 Replicated (n =10) Mite species Replicate Number % with experiments to test whether number recovered coloured gut starved females of three mite species feed on the blood of young chicken (10 days since Dermanyssus galinae 11191 hatching) 21250 3425 4786 51070 61173 71275 81080 91060 10 8 62 Hypoaspis aculeifer 1180 2170 3100 4150 5130 6140 7120 8110 9140 10 18 0 Androlaelaps casalis 1120 2140 3110 Shown are the number of mites 460 recovered after 18 h from the 20 5150 individuals initially released per 680 container and the percentage of 780 these recovered mites with 8140 a red-brown coloured gut. Cli- 9110 mate room conditions were 26°C, 10 6 0 RH >50% and 15 h of darkness two environments and then sometimes in large numbers: Zerconopsis remiger in starling nests and Blattisocius keegani in poultry farms (Table 1). Strikingly, another laelapid spe- cies, Hypoaspis miles (Berlese) (probably Stratiolaelaps scimitus (Womersley); F. Faraji, pers. obs. 2008), currently used to control D. gallinae and other blood feeding mites (e.g., the snake mite, Ophionyssus natricis (Gervais)) on pet animals, such as canaries, pigeons and reptiles (J. Evers, REFONA BV, pers. comm. 2008), was never found spontaneously in starling nests as well as in poultry farms. Only a few individuals of yet another laelapid species, Hypoaspis aculeifer, were found at the end of the breeding season in two starling nests at Vosbergen estate (Eelde-Paterswolde, The Netherlands). This raises the question whether the putative D. gallinae predators found in association with D. gallinae in bird nests under natural conditions oVer perspectives for control of D. gallinae and how they compare to H. miles, currently used in practice for control of D. gallinae on pet animals but not (yet) in poultry farms. This question needs to be answered in future experiments in cages and poultry farms. Below, we discuss the arguments as to why the two Hypoaspis species and one of the four species of putative D. gallinae predators from starling nests, A. casalis, represent candidate predators for biocontrol of D. gallinae in poultry houses. The laelapid mites, H. miles and H. aculeifer, are mainly ground-dwelling predators and they are occasionally reported to occur in nests of birds in low to very low numbers (e.g., Gwiazdowicz et al. 1999; Kristofík et al. 2003; Fenda and Lengyel 2007). Yet, they seem to be somewhat more numerous in nests of ground-nesting birds, such as European bee-eaters 1 C Exp Appl Acarol (2009) 48:63–80 75 and sand martins (Kristofík et al. 1996). We found H. aculeifer in nest boxes with starling nests but then only very late in the breeding season and in very low numbers, even when D. gallinae infestations emerged early in the breeding season. All these observations do not lend support to the hypothesis that H. miles and H. aculeifer have a strong association with bird nests, let alone with D. gallinae infestations. The observations are more consistent with the hypothesis that they inhabit soil litter and visit bird nests. Testing whether H. miles and H. aculeifer in the soil litter exhibit an aggregative response to the density of D. gallinae in nests of ground-nesting birds, has not yet been done, but it seems a feasible and informative experiment to do. Both H. miles and H. aculeifer may well be opportunistic generalist predators of D. gal- linae under natural conditions. We found strong evidence for H. aculeifer females attacking and feeding on all stages of D. gallinae (Table 2; Fig. 3). In fact, starved females of H. aculeifer are voracious predators once released in Petri dishes with D. gallinae and they resume oviposition within a day (I. Lesna, pers. obs.). Also, H. miles has proven to be a voracious predator of D. gallinae and has proven to reproduce on an exclusive diet of D. gallinae (Tuovinen 2008; J. Evers, pers. comm. 2008). For reasons unclear to us, H. miles and H. aculeifer have never been considered to be omnivores that feed on D. galli- nae as well as the blood from the host of D. gallinae. Clearly, the mouthparts of these spe- cies exhibit none of the adaptations known for true blood-feeding acarines. Our blood-feed tests with H. aculeifer and young chicken also did not provide any evidence for feeding on blood of the chicken, even when in a stage where it is most vulnerable to blood-feeding ectoparasites (Table 4). We therefore conclude that H. aculeifer and probably also H. miles are true predators and candidate agents for biocontrol of D. gallinae in poultry houses. The only cautionary remark is that H. aculeifer (in contrast to A. casalis) cannot survive at 32– 34°C in the brood incubator where we carried our blood-feed tests on young hatchlings of European starlings and that such temperatures do occur on hot summer days in Dutch poul- try houses (N. Harteveld, pers. comm. 2008). The laelapid mite, A. casalis, has been reported as a frequent and sometimes abundant inhabitant of the nests from a wide variety of bird species: white storks (Bloszyk et al. 2005), white-tailed sea-eagles (Fenda and Lengyel 2007), owls (Kristofík et al. 2003; Gwiazdowicz 2003), eagles, harriers, buzzards, kites, ospreys, falcons (Gwiazdowicz 2003), house wrens (Pacejka and Thompson 1996; Pacejka et al. 1998), woodpeckers (Pung et al. 2000), Euro- pean bee-eaters (Kristofík et al. 1996), reed warblers (Kristofík et al. 2001) and European starlings (this article). Hence, its common name is ‘the cosmopolitan nest mite’. It has also been found in poultry houses in the UK, Egypt and The Netherlands in considerable num- bers (McKinley 1963; Brady 1970a, b; El-Kammah and Oyoun 2007; this article). Its trophic position in the food web of organisms in bird nests has been unclear. McKin- ley (1963) observed A. casalis feeding on droplets of human blood, but is of the opinion that A. casalis cannot penetrate mammalian or avian skin and can therefore not be a blood parasite, a view shared by Hughes (1976) and Tenquist and Charleston (2001). Indeed, the structure of its gnathosoma and in particular its chelicerae are very diVerent from other blood-sucking mesostigmatic mites, such as Dermanyssidae (Dermanyssus gallinae; McKinley 1963; Roy and Chauve 2007), Macronyssidae (Ornithonyssus sylviarum; Evans 1957) and Rhinonyssidae (Sternostoma tracheacolum Lawrence; Evans 1957). Other authors assume that A. casalis is a (facultative) blood parasite (Men 1959; Radovsky 1985, 1994; Kristofík et al. 1996; Pacejka and Thompson 1996; Pacejka et al. 1996, 1998; Pung et al. 2000; Phillips 2000; Rosen et al. 2002; Svana et al. 2006; Fenda and Lengyel 2007; for review, see Proctor and Owens 2000). However, despite the presence of A. casalis in bird nests in relatively high numbers negative eVects on brood performance have not been 1 C 76 Exp Appl Acarol (2009) 48:63–80 found in the house wren (Pacejka and Thompson 1996; Pacejka et al. 1998) and in the red-cockaded woodpecker (Pung et al. 2000). This may indicate that blood parasitism is not the predominant mode of feeding in A. casalis, or it may even be absent altogether. Some authors consider the possibility that A. casalis is a predator of small arthropods (including mites) in addition to being a blood parasite (Pacejka and Thompson 1996; Pacejka et al. 1996; Kristofík et al. 1996). Using a statistical method called path analysis, Pacejka et al. (1998) found a positive, direct eVect of the numbers of Dermanyssus hirundinis (Hermann) on the numbers of A. casalis. Referring to Pacejka et al. (1998), Proctor and Owens (2000) hypothesize A. casalis to be a predator of blood-feeding mites and therefore a potential mutualist to the nesting bird. Pacejka et al. (1998) did not consider the eVects of arthropods other than blood feeding mites and overlooked the possibility that A. casalis might be a predator of non-parasitic mites as well as young, vulnerable stages of parasitic and non-par- asitic insects in bird nests. For example, A. casalis can feed on several astigmatic mite spe- cies (McKinley 1963) and has been considered as a biocontrol agent of astigmatic mites that are pests in stored products (Barker 1968). Various species of astigmatic mites in stored products are also found in bird nests, where they are probably feeding on fungi. Clearly, the trophic relations of A. casalis with other nest-dwelling arthropods and with the bird are in need of a causal experimental analysis. In this article we provide quantitative evidence supporting the hypothesis that A. casalis is a predator of the poultry red mite (D. gallinae) (Table 2; Fig. 4) and of astigmatic mites living in nests (Glycyphagus domesticus (De Geer), Lepidoglyphus destructor (Schrank)) and in poultry houses (Tyrophagus putrescentiae). We also show quantitatively that A. casalis could not extract blood from young chicken and young starlings (Tables 3, 4), i.e., from birds in a life stage where they are most accessible and vulnerable to acarine blood parasites. These results are largely in agreement with the qualitative observations reported by McKinley (1963). This author also observed starved A. casalis feeding on astigmatic mites and poultry red mites and found no evidence (based on mite gut colour- ation or bite marks on host skin) for feeding on chicken, laboratory mice and men (even when the human skin was treated with Wne sandpaper to improve access to blood vessels). Thus, given its numerical abundance in starling nests, its inverse association with D. galli- nae in starling nests, its spontaneous occurrence in poultry houses infested with D. galli- nae, its ability to complete its life cycle on a diet of D. gallinae and its inability to acquire blood directly from a bird (even when in its most vulnerable stage), we conclude that A. casalis is a true predator. This conclusion should challenge ornithologists to revise their views on how to judge (and perhaps how birds judge) the risk of ectoparasitism in a nest and their views on why birds re-use old nesting sites or old nest material (e.g., Mazgajski (2007) considers old nest material of European starlings only as a source of ectoparasites). Moreover, it points at a new candidate agent for the control of D. gallinae in poultry houses. Apart from H. miles, H. aculeifer and A. casalis, there may well be several other candi- date predators for biocontrol of D. gallinae. Our inventory of mites in starling nests and poultry farms in The Netherlands yielded species that may act as predators of D. gallinae, like C. eruditus, Z. remiger and B. keegani. Also, our attempt to review the literature on mites in bird nests appeared to provide several nidicolous species, such as Hypoaspis lubrica Voigts and Oudemans (Brady 1970a, b; Gwiazdowicz et al. 1999). However, none of these species have been actually tested for their ability to feed on D. gallinae, nor for their (in-)ability to feed on the blood of their hosts. Some of these species have been observed to feed on free blood (e.g., H. lubrica feeding on free blood of mice; Li and Meng 1992), but as shown for the case of A. casalis feeding on free blood is not suYcient evidence to infer haematophagy (McKinley 1963; this article). 1 C Exp Appl Acarol (2009) 48:63–80 77
As much as we realize that our conclusions on the exclusive predatory life style of A. casalis and H. aculeifer may not hold under all conditions (e.g., other bird species), we caution against uncritical citing and unconWrmed inferences on feeding life styles of nidicolous arthropods (e.g., Pacejka et al. 1998; Xing-Yuan et al. 2007). Bird nests may not only harbour ectoparasites and fungivores but also predators of one or even both of these trophic guilds. Indeed, the arthropods in nests are part of a multitrophic system.
Acknowledgments We (IL, MWS) thank René Heijnen (CeHaVe, Veghel, The Netherlands) and Karel Bolckmans (Koppert BV, Berkel en Rodenrijs, The Netherlands) for stimulating us to work on biological con- trol of poultry red mites. Thea Fiks-Niekerk, Monique Mul and Rick van Emous (Animal Sciences group, Wageningen University, Lelystad, The Netherlands) were instrumental in providing information on pest sta- tus of poultry red mites and current practice of poultry management. Sampling of the mite fauna of poultry houses (4 farms in the province of Brabant and Limburg, The Netherlands) was enabled with much help of Nico Harteveld (Koppert BV, Berkel en Rodenrijs, The Netherlands) and Sieds van der Schaaf (Protekta, Gemert, The Netherlands). Ria Wiltenburg and Marnix van de Wetering (CeHaVe, Veghel, The Netherlands) are thanked for their help in realizing the required WOD licence (Wet op de Dierproeven) for tests on blood feeding by mites on young chicken, experiments that were carried out at the research farm ‘Laverdonk’ (Ce- HaVe, Heeswijk, The Netherlands). Two M.Sc. students of the University of Groningen, Judith J. Westveer and Linda M. van Zomeren, are thanked for their help with tests on blood feeding by mites on young starlings, experiments that were carried out at Vosbergen estate (Paterswolde, The Netherlands) under WOD licences held by Prof. Jan Komdeur (University of Groningen, Haren, The Netherlands). Tuomi Tuovinen (MTT-Ag- rifood Research, Jokioinen, Finland) and John Evers (REFONA BV, Westerbork, The Netherlands) are thanked for sharing insights on the use of Hypoaspis miles for poultry red mite control. Prof. Urs Wyss (Uni- versity of Kiel, Germany) is thanked for his skills, patience and help to make video-records of mite behaviour. The work presented in this article was enabled by Wnancial support from the Netherlands Academy of Sci- ences (KNAW, Amsterdam, The Netherlands), the Dutch Ministry of Agriculture, Nature and Food Quality (LNV, Expertise Center, Bennekom, The Netherlands) and the Product Boards for Livestock, Meat and Eggs (PVE, Zoetermeer, The Netherlands).
Appendix
To test for conspeciWcity of Dermanyssus samples from starling nests and those from poultry farms, three gene regions were sequenced and aligned with homologous sequences from pop- ulations of various origins and diverse Dermanyssus species: (1) a fragment of 18S–28S rRNA, including internal transcribed spacer 1 (ITS1), 5.8S rRNA and ITS2 (nuclear gene region) (2) 16S rRNA (mitochondrial gene region), and (3) mt-COI (a mitochondrial protein- coding region of cytochrome oxidase subunit I). Populations collected in The Netherlands are labelled by an acronym (IL) referring to the Wrst author of this article, followed by the number of the starling’s nest box at Vosbergen estate (Eelde-Paterswolde, The Netherlands). The EMBL Accession numbers are provided below to enable citation of database entries.
Gene region Population EMBL accession number
ITS IL213 FM207490 ITS IL227 FM207491 16S rRNA IL213 FM207492 16S rRNA IL227 FM207494 mt-COI IL302 FM207495 mt-COI IL227 FM207496 mt-COI IL202A FM207497 mt-COI IL202C FM207498 mt-COI IL213 FM207499
1 C 78 Exp Appl Acarol (2009) 48:63–80
On the basis of mt-COI, Roy et al. (2009) show that four populations (IL302, IL227, IL202, IL213) sampled from starling nests belong to D. gallinae. Roy et al. (2008) show that the species delineation apparent from the mt-COI tree is conWrmed by other analyses, which include all three gene regions (two mitochondrial and one nuclear). Because the samples from starling populations branch within the most distal D. gallinae clade (and nei- ther in a sister clade, nor at the basis of the large D. gallinae clade), we infer that they are conspeciWc. Most likely, all 29 populations from starling-occupied nest boxes at Vosbergen estate in 2007 are conspeciWc, because Roy et al. (2009) found strong evidence for single or very similar haplotypes of Dermanyssus per bird host and geographical location. Finally, Roy et al. (2009) show that in contrast to others the clade D. gallinae groups together all populations represented in poultry farms (layer hens) or breeding facilities for canaries, other Fringillidae and pigeons. Thus, the clade D. gallinae harbours synanthropic, as well as bird-associated populations.
References
Abo-Taka SM (1996) Mites inhabiting poultry farms in Egypt. In: Mitchell R, Horn DJ, Needham GR, Welbo- urn WC (eds) Acarology IX, proceedings, vol 1. Ohio Biological Survey, Columbus, pp 97–99 Axtell RC, Arends JJ (1990) Ecology and management of arthropods pests of poultry. Annu Rev Entomol 35:101–126. doi:10.1146/annurev.en.35.010190.000533 Barker PS (1968) Bionomics of Androlaelaps casalis (Berlese) (Acarina: Laelapidae), a predator of mite pests of stored cereals. Can J Zool 46:1099–1102. doi:10.1139/z68-157 Beck W (1999) Farm animals as disease vectors of parasitic epizoonoses and zoophilic dermatophytes and their importance in dermatology. Hautarzt 50:621–628. doi:10.1007/s001050050971 Bloszyk J, Gwiazdowicz DJ, Bajerlein D, Halliday RB (2005) Nests of the white stork Ciconia ciconia (L.) as a habitat for mesostigmatic mites (Acari, Mesostigmata). Acta Parasitol 50(2):171–175 Brady J (1970a) The mites of poultry litter: observations on the bionomics of common species, with a species list for England and Wales. J Appl Ecol 7:331–348. doi:10.2307/2401384 Brady J (1970b) Litter mites and their eVects on poultry. Worlds Poult Sci J 26:658–668. doi:10.1079/ WPS19700022 Brouwer L, Komdeur J (2004) Green nesting material has a function in mate attraction in the European star- ling. Anim Behav 67:539–548. doi:10.1016/j.anbehav.2003.07.005 BuVoni G, Di Cola G, Baumgärtner J, Maurer V (1995) A mathematical model for trophic interactions in an acarine predator–prey system. J Biol Syst 3:303–312. doi:10.1142/S0218339095000289 BuVoni G, Di Cola G, Baumgärtner J, Maurer V (1997) The local dynamics of acarine predator-prey (Cheyletus eruditus–Dermanyssus gallinae) populations: identiWcation of a lumped parameter model. Mitt Schweiz Entomol Ges 70:345–359 Burtt EH Jr, Chow W, Babitt GA (1991) Occurrence and demography of mites of tree swallow, house wren, and eastern bluebird boxes. In: Loye JE, Zuk M (eds) Bird-parasite interactions: ecology, evolution and behaviour. Oxford University Press, Oxford, pp 104–122 (ornithology Series) Clark L (1991) The nest-protection hypothesis: the adaptive use of plant secondary compounds by European starlings. In: Loye JE, Zuk M (eds) Bird–parasite interactions: ecology, evolution and behaviour. Oxford University Press, Oxford, pp 205–221 (ornithology Series) El-Kammah KM, Oyoun LMI (2007) Taxonomic records of mite species associated with animal and poultry farms in the Nile Delta, valley and N. Sinai Egypt. In: Morales-Malacara JB, Behan-Pelletier V, Uecker- mann E, Perez-T, Estrada-Venegas EG, Badii M (eds) Acarology XI, proceedings of the international congress, pp 223–232 Emous RA, Fiks-Van Niekerk TGCM, Mul MF (2005) Enquete vogelmijten op leghennenbedrijven: D 11. 000.000 schade voor de sector. Pluimveehouderij 35(22 januari):8–9 Evans GO (1957) An introduction to the British Mesostigmata (Acarina) with keys to families and genera. Zool J Linn Soc 43:203–259. doi:10.1111/j.1096-3642.1957.tb01552.x Faraji F, Bakker FM (2008) A modiWed method for clearing, staining and mounting plant-inhabiting mites. Eur J Entomol 105:793–795 Fenda P, Lengyel J (2007) Rostoce (Acarina: Mesostigmata) v hniezdach orliaka morskeho (Haliaeetus albicil- la) na Slovensku. Entomofauna carpathica 19:48–50
1 C Exp Appl Acarol (2009) 48:63–80 79
Gupta SK, Paul K (1989) Nest-associated acarines of birds in India. In: Channabasavanna GP, Viraktamath CA (eds) Progress in acarology, vol 2. EJ Brill, Leiden, The Netherlands, pp 315–321 Gwiazdowicz DJ (2003) Mites (Acari: Mesostigmata) occurring in the nests of birds of prey (Falconiformes) and owls (Strigiformes). Acarina 11:235–239 Gwiazdowicz DJ, Mizera T, Skorupski M (1999) Mites from greater spotted eagle nests. J Raptor Res 33:257–260 Gwinner H, Berger S (2005) European starlings: nestling condition, parasites and green nest material during the breeding season. J Ornithol 146:365–371. doi:10.1007/s10336-005-0012-x Hughes AM (1976) The Mites of stored food and houses. Second Edition, Ministry of Agriculture, Fisheries and Food, Technical Bulletin 9:1–400 Kristofík J, Masan P, Sustek Z (1996) Ectoparasites of bee-eater (Merops apiaster) and arthropods in its nests. Biologia 51:557–570 Kristofík J, Masan P, Sustek Z (2001) Mites (Acari), beetles (Coleoptera) and Xeas (Siphonaptera) in the nests of great reed warbler (Acrocephalus arundinaceus) and reed warbler (A. scirpaceus). Biologia 56:525–536 Kristofík J, Masan P, Sustek Z, Kloubec B (2003) Arthropods (Pseudoscorpionida, Acari, Coleoptera, Sipho- naptera) in nests of the Tengmalm’s owl, Aegolius funereus. Biologia 58:231–240 Lesna I, Sabelis MW, Bolland HR, Conijn CGM (1995) Candidate natural enemies for control of Rhizogly- phus robini Claparède (Acari: Astigmata) in lily bulbs: exploration in the Weld and preselection in the laboratory. Exp Appl Acarol 19:655–669. doi:10.1007/BF00145254 Li G, Meng Y (1992) Investigation in sucking mice free blood habit of Hypoaspis lubrica. Chin J Vector Biol Control 3:9–12 (in Chinese) Lundqvist L (1995) Ectoparasitic mites in the nest of starlings (Sturnus vulgaris); do they play a role in the sexual selection of the birds? In: Kropczynska D, Boczek J, Tomczyk A (eds) The Acari: physiological and ecological aspects of Acari-host relationships. OWcyna Dabor, Warszawa, Poland, pp 545–548 Maurer V, Baumgärtner J (1992) Temperature inXuence on life table statistics of the chicken mite Dermanys- sus gallinae (Acari: Dermanyssidae). Exp Appl Acarol 15:27–40. doi:10.1007/BF01193965 Maurer V, Baumgärtner J (1994) A population model for Dermanyssus gallinae (Acari, Dermanyssidae). Exp Appl Acarol 18:409–422. doi:10.1007/BF00051523 Maurer V, Hertzberg H (2001) Ökologische legehennenhaltung. Was tun gegen die kleinen Vampire? DGS- Magazin. Woche 40:49–52 Maurer V, Baumgärtner J, Bieri M, Fölsch DW (1993) The occurrence of the chicken mite Dermanyssus gallinae (Acari: Dermanyssidae) in Swiss poultry houses. Mitt Schweiz Entomol Ges 66:87–97 Mazgajski TD (2007) EVect of old nest material in nest boxes on ectoparasite abundance and reproductive output in the European starling Sturnus vulgaris (L.). Pol J Ecol 55:377–385 McKinley DJ (1963) The morphology and biology of Haemolaelaps casalis Berlese (Acarina: Mesostig- mata). Ann Mag Nat Hist 13(6):65–76 Men YT (1959) Concerning the feeding of the mite Haemolaelaps casalis (Gamasoidea, Parasitiformes). Medskaya Parazitol 28:603–609 (in Russian) Nordenfors H, Chirico J (2001) Evaluation of a Sampling Trap for Dermanyssus gallinae (Acari: Derman- yssidae). J Econ Entomol 94:1617–1621 Pacejka AJ, Thompson CF (1996) Does removal of old nests from nest boxes by researchers aVect mite populations in subsequent nests of house wrens? J Field Ornithol 67:558–564 Pacejka AJ, Santana E, Harper RJ, Thompson CF (1996) House wrens Troglodytes aedon and nest-dwelling ectoparasites: mite population growth and feeding patterns. J Avian Biol 27:273–278. doi:10.2307/ 3677258 Pacejka AJ, Gratton CM, Thompson CF (1998) Do potentially virulent mites aVect house wren (Troglodytes aedon) reproductive success? Ecology 79:1797–1806 Philips JR (2000) Review and checklist of the parasitic mites (Acarina) of the Falconiformes and Strigifor- mes. J Raptor Res 34:210–231 Philips JR, Dindal DL (1979) The acarine community of nests of birds of prey. In: Rodriguez JG (ed) Recent advances of acarology, vol 1. Academic Press, New York, pp 559–562 Philips JR, Poole A, Holt D (1989) Nest mites of ospreys and short-eared owls in Massachusetts salt marshes, USA. In: Channabasavanna GP, Viraktamath CA (eds) Progress in acarology, vol 2. EJ Brill, Leiden, The Netherlands, pp 305–307 Proctor H, Owens I (2000) Mites and birds: diversity, parasitism and coevolution. Trends Ecol Evol 15:358–364. doi:10.1016/S0169-5347(00)01924-8 Pung OJ, Carlile LD, Whitlock J, Vives SP, Durden LA, Spadgenske E (2000) Survey and host Wtness eVects of red-cockaded woodpecker blood parasites and nest cavity arthropods. J Parasitol 86:506–510
1 C 80 Exp Appl Acarol (2009) 48:63–80
Putatunda BN, Gupta SK, Singh J (1989) Acarine associates of birds in West Bengal, India. In: Channaba- savanna GP, Viraktamath CA (eds) Progress in acarology, vol 2. EJ Brill, Leiden, The Netherlands, pp 309–313 Quintero MT, Acevedo AC (1984) Studies on deep litter mites in poultry farms in Mexico. In: GriYths DA, Bowman CE (eds) Acarology VI, vol 1. Ellis Horwood Ltd, Chichester, pp 629–634 Rabash J, Browne W, Healy M, Cameron B, Charlton C (2008) “MLwiN v2.10 beta 5” Centre of Multilevel Modelling. University of Bristol, UK Radovsky FJ (1985) Evolution of mammalian mesostigmate mites. In: Kim KC (ed) Coevolution of parasitic arthropods and mammals. Wiley, New York, pp 441–504 Radovsky FJ (1994) The evolution of parasitism and the distribution of some dermanyssoid mites (Mesostig- mata) on vertebrate hosts. In: Houck MA (ed) Mites: ecological and evolutionary analysis of life history patterns. Chapman and Hall, New York, pp 186–217 Rosen S, Yeruham I, Braverman Y (2002) Dermatitis in humans associated with the mites Pyemotes tritici, Dermanyssus gallinae, Ornithonyssus bacoti and Androlaelaps casalis in Israel. Med Vet Entomol 16:442–444. doi:10.1046/j.1365-2915.2002.00386.x Roy L, Chauve CM (2007) Historical review of the genus Dermanyssus Dugès (Acari: Mesostigmata: Der- manyssidae). Parasite 14:87–100 Roy L, Dowling APG, Chauve CM, Buronfosse T (2008) Delimiting species boundaries within Dermanyssus Dugès, 1834 (Acari: Mesostigmata) using a total evidence approach. Mol Phylogenet Evol (in press). doi:10.1016/j.ympev.2008.11.012 Roy L, Dowling APG, Chauve CM, Lesna I, Sabelis MW, Buronfosse T (2009) Molecular phylogenetic assessment of host range in Dermanyssus species. Exp Appl Acarol. doi:10.1007/s10493-008-9231-1 Svana M, Fenda P, Orszaghova Z (2006) Rostoce (Acarina: Mesostigmata) v hniezdach vtakov jz Slovenska. Folia faunistica Slovaka 11:39–42 Tenquist JD, Charleston WAG (2001) A revision of the annotated checklist of ectoparasites of terrestrial mammals in New Zealand. J R Soc N Z 31:481–542 Tuovinen T (2008) Predatory arthropods as biocontrol agents of Dermanyssus gallinae? In: BSP Spring, trypanosomiasis/leishmaniasis & malaria meetings. Newcastle, UK, March 30–April 2. Abstract booklet: 170 Valiente Moro C, Chauve C, Zenner L (2005) Vectorial role of some dermanyssoid mites (Acari, Mesostig- mata, Dermanyssoidea). Parasite 12:99–109 Valiente Moro C, Chauve C, Zenner L (2007) Experimental infection of Salmonella Enteritidis by the poultry red mite, Dermanyssus gallinae. Vet Parasitol 146:329–336. doi:10.1016/j.vetpar.2007.02.024 Veiga JP, Polo V, Vinuela J (2006) Nest green plants as a male status signal and courtship display in the spot- less starling. Ethology 112:196–204. doi:10.1111/j.1439-0310.2006.01148.x Xing-Yuan M, Xian-Guo G, Wen-Ge D, Ai-Qin N, Ti-Jun Q, Dian W (2007) Ectoparasites of Chevrier ‘s Weld mouse, Apodemus chevrieri, in a focus of plague in Southwest China. Med Vet Entomol 21:297–300
1 C 10.3 Annexe 3 : numéros d'accès des séquences obtenues d'EF1-alpha et amorces correspondantes
Espèce Isolat Numéro d'accès EMBL Androlaelaps casalis 2.4 AM930875 Androlaelaps casalis ACA AM930874 Dermanyssus apodis GO15 AM930867 Dermanyssus apodis GO16 AM930866 Dermanyssus apodis MAR AM930870 Dermanyssus carpathicus 5 AM930871 Dermanyssus carpathicus RQ AM930872 Dermanyssus gallinae CANIT AM930877 Dermanyssus gallinae Chab AM930857 Dermanyssus gallinae DR AM930856 Dermanyssus gallinae Fa1 AM930855 Dermanyssus gallinae GO12 AM930865 Dermanyssus gallinae GO26 AM930869 Dermanyssus gallinae GO44 AM930868 Dermanyssus gallinae LB AM930876 Dermanyssus gallinae PO1 AM930861 Dermanyssus gallinae SK AM930858 Dermanyssus gallinae Woodp AM930862 Dermanyssus gallinae (L1) COL AM930854 Dermanyssus gallinae (L1) GO8 AM930864 Dermanyssus gallinae (L1) LC AM930863 Dermanyssus hirundinis HR AM930859 Dermanyssus hirundinis OC AM930860 Dermanyssus longipes PAS AM930873 Ornithonyssus sylviarum JBO105 AM930881
Amorces pour l’obtention des séquences d’EF1-alpha :
DgEF1-F 5' TGGGCAAGGGCTCCTTCAAGTA 3' DgEF1-R 5' TCGCACTTCTCCTTAATCTCCTTGAA 3' AcEF1-F 5' CTGTGGAAGTTCGAGACGCC 3' AcEF1-R 5' CTCGTGGTGCATTTCGACCGACTTC 3'
Le couple DgEF1-F + DgEF1-R génère un amplicon d’env. 1000 pb, généralement faiblement amplifié. Le couple AcEF1-F + AcEF1-R génère un amplicon un peu plus court (env. 900 pb), la réaction de PCR est en général beaucoup plus performante (intensité de la bande et succès du séquençage nettement accrus).
5 Même séquence exactement que PM (3 individus à chaque fois), un autre isolat d’O. sylviarum, et que D. gallinae (lignes précédentes). Provenances des isolats d’O. sylviarum : PM, Vaulx-en-Velin, Rhône, France ; JBO10, Mollégès, Bouches-du-Rhône, France.
296